US inventories have risen to an elevated level relative to sales, suggesting that the 3-5 year Kitchin stockbuilding cycle is about to turn down. Together with slowing real narrow money expansion, this casts doubt on 2016 growth prospects.

A post in August 2014 argued that US economic fluctuations can be explained by the interaction of three cycles: the 3-5 year Kitchin stockbuilding cycle; the 7-11 year Juglar business investment cycle; and the 15-25 year Kuznets housebuilding cycle. Severe recessions, such as 2008-09 and 1974-75, occur when the three cycles reach a trough in the same year.

The post suggested that the Kitchin cycle last hit bottom in 2012. GDP grew by just 1.3% in the year to the fourth quarter of 2012, with economic weakness contributing to the Fed’s decision to launch QE3 in September. If this dating is correct, the cycle is scheduled to reach another low between 2015 and 2017.

A key indicator of the status of the Kitchin cycle is the ratio of inventories to final sales. The ratio typically moves significantly above its trend at cycle peaks, falling beneath it at troughs. It was above trend in the first and second quarters of 2015, with the deviation larger than at the last cycle peak in 2010 – see first chart.

A return of the ratio to trend requires a cut in the stock of inventories and / or a rise in final sales. With real money expansion slowing, final sales are unlikely to surge. Suppose that they grow by 2.5% over the coming year. The stock of inventories would then need to be held stable to return the ratio to sales to trend in 12 months’ time.

Inventories, however, were still rising fast in the second quarter – by the equivalent of 0.7% of GDP. A fall in the rate of change to zero, therefore, would imply a drag on GDP of this magnitude.

If the inventories to sales ratio were to fall beneath trend, as usually occurs at Kitchin cycle troughs, the implied drag would, of course, be larger.

As noted in the previous post, Kitchin cycle downturns, in isolation, are not usually associated with recessions. The Kuznets housebuilding cycle remains in an upswing. The Juglar business investment cycle is scheduled to reach another low between 2016 and 2020. Juglar downturns are almost always preceded by a profits squeeze. Economic profits* have stagnated over the past year but are not weak enough to suggest a fall in investment – second chart.

With prospects for business spending softening, personal consumption and housebuilding will need to make stronger contributions if GDP growth is to be sustained at its recent pace.

*Economic profits adjust for stock appreciation and over / underreporting of depreciation.

Eurozone money measures rose strongly again in July, reinforcing the positive view here of economic prospects. Narrow money is growing much faster than in the US, suggesting superior Eurozone economic performance over the next six to 12 months.

Broad money M3 and narrow money M1 rose by 0.8% and 1.2% respectively in July. Annual M3 growth was 5.3%, equal to a high reached in April, while M1 growth surged to 12.1% – the strongest since 2009.

The money measure with the best economic forecasting properties, according to ECB studies, is non-financial M1, comprising currency and overnight deposits held by households and non-financial corporations. Six-month growth in this measure has been running at about 5% since late 2014, with July at 5.2%, or 10.7% annualised – see first chart.

Growth in real terms (i.e. relative to consumer prices) has moderated since end-2014, reflecting a recovery in inflation. However, the current six-month increase of 4.3%, or 8.9% annualised, remains strong by historical standards, suggesting solid economic prospects. Six-month inflation, moreover, is likely to pull back near term on oil and exchange rate effects.

In addition to money supply strength, bank loans to the private sector, adjusted for sales and securitisation, rose by 0.3% in July, pushing annual growth up to 1.4% – the fastest since 2011. Unlike money, credit is a coincident or lagging indicator of the economy. With narrow money trends signalling solid economic growth, lending should pick up further.

US real narrow money M1 rose by 2.1%, or 4.2% annualised, in the six months to July*. The recent gap between Eurozone real non-financial M1 growth and US real M1 growth is the largest since early 2010. GDP grew faster in the Eurozone than the US over the subsequent year**.

The ECB publishes a country breakdown of overnight deposits but not currency. Among the big four, six-month growth of real deposits is strongest in France and Spain, with Italy lagging – second chart. Consistent with the positive French signal, Insee business surveys for August were upbeat, with composite climate indicators for services and industry at their highest since 2011.

The outflow of deposits from Greek banks slowed to €1.7 billion in July after €8.1 billion in June but, surprisingly, the overnight component rebounded, reducing the six-month decline in real deposits sharply – third chart. Growth remains strong in Ireland but fell back in Portugal after a spike last month.

*US non-financial M1 is available only quarterly and with a lag – the latest datapoint is for end-March.
**GDP rose by 2.8% and 1.9% respectively in the year to the first quarter of 2011.

UK household spending is as extended relative to income as in the late 1980s and mid 2000s, ahead of the 1990-91 and 2008-09 recessions. Those earlier episodes were associated with rapid and unsustainable credit growth; current spending excess, by contrast, is the mirror of record-low accumulation of financial assets. A normalisation of financial investment behaviour threatens MPC and OBR central forecasts that GDP growth will be sustained at about 2.5% per annum over the next three to five years.

The suggestion that household finances are weak runs counter to recent commentary noting stronger nominal and real wage growth and a significant decline in the debt / income ratio from its 2009 peak. Spending, however, has accelerated along with income, while lower gearing has been achieved mainly by paring asset accumulation.

The difference between income and spending is equal to “net lending” in the capital account of the national accounts. Household sector net lending was negative by £5.4 billion, or 1.7% of available income (resources), in the first quarter of 2015. In data extending back to 1987, deficits were previously recorded only in 1988, 1999 and over 2004-08 – see first chart.



The weakness of household finances in the late 1980s and mid 2000s was transmitted to spending as interest rates subsequently rose. Rates, by contrast, fell in 1999, while the trend in net lending – as measured by a four-quarter moving sum – did not fall as far as in the other episodes. Growth slowed in the early 2000s but remained respectable. With interest rates likely to rise soon, and net lending in persistent deficit, current conditions bear a closer resemblance to the late 1980s and mid 2000s.

Household net lending should be distinguished from “saving”, which is the difference between income and consumption. Saving, in other words, includes household spending on investment goods – mainly housing. The saving ratio – saving as percentage of income – was 4.9% in the first quarter of 2015. Subtracting investment net of capital transfers of 6.6% of income generates the net lending deficit of 1.7% of income. The saving ratio is low by historical standards – it was below the current level in only three quarters in data extending back to 1963.

While net lending is similar now to the late 1980s and mid 2000s, borrowing / saving behaviour is very different. The net lending position in the capital account is mirrored in the financial account by the difference between gross lending or financial investment (“net acquisition of financial assets”) and borrowing (“net acquisition of financial liabilities”). This alternative financial account measure is not exactly equal to net lending in the capital account because of a statistical discrepancy, which is usually positive – second chart. Assuming that the capital account is more accurate, this implies that the financial account either overstates financial investment or understates borrowing.



The weakness of net lending in the late 1980s and mid 2000s, and to a lesser extent 1999, reflected strong borrowing, which peaked at 16.1% and 17.1% of income respectively in the former two cases (four-quarter moving sum) – third chart. Despite a recent recovery, borrowing currently remains historically very subdued, at 3.4% of income.

Weak net lending today is explained, instead, by low financial asset accumulation – currently only 4.0% of income versus a pre-crisis average (1987-2007) of 16.1%.

 

Households have slashed their asset accumulation probably for two main reasons, both related to monetary policy. First, low interest rates have reduced the demand for bank deposits and other money-like assets. This has been reflected in monetary data showing a persistent withdrawal of funds from household time deposits – the outflow totalled £23 billion in the 12 months to June.

Secondly, super-loose monetary policy has inflated the valuations of existing equity and bond holdings, allowing households to scale back their acquisition of new assets while maintaining a rise in wealth – fourth chart. Asset price inflation, in other words, has temporarily reduced the need for financial saving.



Household spending turned down in the late 1980s and mid 2000s as rising interest rates and other developments triggered a normalisation of borrowing behaviour. The risk now is that households will reduce spending to return asset accumulation to a more normal historical level.

Such a scenario would probably involve a rise in interest rates but could occur simply as a result of asset price inflation slowing. With smaller price gains, households would need to raise their acquisition of assets to maintain growth in financial wealth. (This is analogous to the “credit impulse” idea that spending is related to the rate of change of borrowing rather than its level.)

To summarise, it is likely that households will wish to return their financial asset accumulation / net lending to a more normal historical level at some point over the next few years. Barring a substantial rise in income, this would require spending restraint, with negative implications for economic growth.

Could the adjustment be achieved solely via income expansion? This would require a rise in spending relative to income in other sectors of the economy. The equality of aggregate spending and income implies the following identity for the net lending positions of the various sectors:

Household + corporate + government + rest of world = 0

Since the net lending position of the rest of the world is, to a close approximation, the mirror-image of the current account balance, this is equivalent to:

Household + corporate + government = current account

A rise in household net lending would require the corporate / government positions to deteriorate and / or the current account to improve. The Chancellor’s plans, of course, involve eliminating the government deficit. An income-driven recovery in household net lending, therefore, would depend on big rise in corporate spending relative to income and / or a surge in net overseas earnings.

Corporate net lending has moved from a large surplus after the 2008-09 recession to a modest deficit – fifth chart. Scarred by the crisis, and uncertain about the financial / economic impact of monetary policy normalisation, firms may be reluctant to allow a significantly larger shortfall.



The current account position may improve as the recent puzzling weakness of net investment earnings reverses but there is little reason to expect significantly stronger export performance, with the most dynamic sector – financial services – now hobbled by overregulation.

The prospective combined change in the corporate and rest of world positions, in other words, is unlikely to be sufficient to accommodate both government deficit reduction and a pain-free recovery in household net lending.

Current negative sentiment in markets reflects perceived US / Chinese economic divergence. Chinese weakness is widely thought to be intensifying, dragging down global growth, while the US economy is deemed sufficiently solid to keep the Fed on track for an early rate hike. The suggestion is that equity investors face an unappealing combination of softening corporate earnings prospects and a higher discount rate (with real Treasury yields rising recently despite falling nominal rates).

The combination of Chinese economic weakness and US solidity at mid-year had been suggested by narrow money trends at the start of 2015. Six-month growth of US real M1 was rising strongly while the corresponding Chinese measure* had fallen to its lowest level in data extending back to 2000 – see first chart.

Current trends, however, are giving an opposite message. Chinese real M1 growth has revived, with recent policy easing suggesting a further lift, while US growth has fallen sharply, dropping below the Chinese level in July for the first time since 2013.

This suggests that Chinese economic growth will recover by end-2015, if not before, while US strength will fade. A US slowdown may arrive too late to prevent a Fed rate hike but would be likely to push back any follow-up move.

The combination of Chinese economic revival with a Fed back on hold would probably be viewed positively by equity investors.

The risk, of course, is that the US economy slows “too much”, resulting in earnings concerns outweighing relief at Fed tightening being pushed back. To rule out this scenario, it is important that US real money growth stabilises or recovers. Inflation trends should be helpful near term, with recent further commodity price falls suppressing headline consumer prices.

The view here remains that current global economic weakness is being exaggerated and that underlying momentum has recovered since the late spring. While today’s Chinese Markit / Caixin manufacturing PMI survey for August was softer again, Japanese and Eurozone readings strengthened – second chart. Yesterday’s US Philadelphia Fed survey was also solid**. The Chinese Markit PMI undershot the official PMI measure in July and may underrepresent larger companies benefiting from stimulus measures; by contrast, the Market News International business survey improved sharply this month.

*Using ”true” M1, including household as well as corporate demand deposits – see previous post.
**The New York Fed survey was notably weak but this may reflect a large fall in the Canadian dollar, which is an important influence for some state manufacturers.

Chinese monetary trends suggest an improving economic outlook. With recent easing measures yet to have their full impact, money growth may lift further during the second half.

The preferred monetary aggregate here is the narrow “true M1” measure, comprising currency in circulation and demand deposits of corporations and households. As previously discussed, the official M1 aggregate omits household demand deposits, making it less reliable as a leading indicator, particularly when the outlook for consumer spending is shifting.

Six-month growth of true M1 rose to 4.8% in July (9.9% annualised), the fastest since December 2013 – see first chart. The increase has been smaller in real terms (i.e. relative to consumer prices) but growth is nonetheless above its average over the past three years.

The six-month change in real true M1 bottomed in December 2014, recovering in early 2015 – second chart. Six-month industrial output growth reached a trough in March, moving higher through July. With real true M1 still accelerating, output growth may remain around the current level or rise further into the autumn.

True M1 expansion picked up strongly in 2008-09 and 2012 after reserve requirements and interest rates were cut – third chart. Recent reductions have been larger than in 2012. Monetary trends have already responded to these actions but a further positive impact is likely.

UK labour market tightening stalled in early 2015: employment and aggregate hours worked fell by 0.2% between the first and second quarters, while the unemployment rate edged up from 5.5% to 5.6%.

The issue for the MPC is whether this setback is temporary, reflecting slower first-quarter GDP expansion and pre-election hiring caution, or indicates a weaker growth trend and / or stepped-up productivity performance.

Vacancies, a leading indicator of employment, provide tentative support for the former hypothesis. The official vacancies series (a three-month moving average) declined between February and May but recovered in June – see first chart. Similarly, claimant count unemployment – which now includes those transitioned to universal credit – fell by 4,900 in July, the most since April.

Annual weekly earnings growth (three-month average) declined from 3.2% in May to 2.4% in June, reflecting a slowdown in volatile bonuses. The ex bonus measure was stable at 2.8%, with the private sector component at 3.3%. With average weekly hours down slightly from a year ago, hourly ex bonus earnings growth rose to 3.1%, a four-year high – second chart.