Previous posts argued that QE programmes in the US, UK and Japan gave little boost to broad money growth, helping to explain their disappointing economic impact. Early indications are that the current Eurozone programme is proving similarly ineffectual.

Monthly growth of broad money M3 averaged 0.4% over March-May, unchanged from the previous six months (QE started in early March).

The earlier posts argued that the QE impact on broad money was small for two main reasons. First, central bank bond purchases were partly offset by sales, or reduced buying, by banks. This demand reduction was a direct result of QE – banks had less need to hold liquid securities because QE was boosting their reserves. Secondly, QE encouraged capital outflows – some of the money created, in other words, flowed abroad.

These effects are evident in the latest Eurozone data. Banks sold €25 billion of government securities over March-May versus ECB purchases of €147 billion. By contrast, banks bought €40 billion in the six months before QE started.

The first chart shows six-month M3 growth and the contributions of the various credit counterparts. The “government contribution” shows the net effect of ECB and bank transactions in government securities (as well as changes in direct lending and government deposits). This is positive but lower than in December / January. Banks’ net external assets, meanwhile, are now acting as a drag on broad money growth, having been strongly expansionary in 2013-14. This implies that the (non-bank) capital account of the balance of payments has moved into substantial deficit, offsetting the large current account surplus.

The biggest contributor to M3 growth recently has been a reduction in bank’s longer-term funding. Holders of maturing bank bonds have been reluctant to reinvest because of the paltry yields on offer. The fall in yields probably reflects the ECB’s interest rate cuts last year and global trends, rather than QE.

While QE has failed to provide an additional boost, monetary trends continue to suggest respectable economic prospects. Six-month growth of real M3 and M1, however, has fallen back as inflation has recovered – second chart.

The country breakdown shows a further rise in growth of real M1 deposits in core countries but a decline in the periphery – third chart. There was a notable slowdown in Italy, while growth rose in Germany and remains strong in Spain and France, though lower than last month – fourth chart.

• Financial crises in peripheral / emerging economies are often symptomatic of tightening global liquidity conditions.

• However, Greece has lacked market access for many months and its crisis reflects the government’s voluntary decision to reject an EU / IMF support programme.

• The global real narrow money growth measure tracked here has slowed recently but remains solid and well ahead of industrial output expansion, implying “excess” liquidity.

• The US and Chinese economies appear to be regaining momentum after weakness in early 2015. The Eurozone is growing moderately while Japan has slowed after a strong first quarter.

• Greece’s recession will now deepen significantly and capital controls may remain in place for months.

• The direct economic and financial implications for the rest of Europe are negligible, reflecting the small size of the Greek economy and the shift of financial exposures to official creditors.

• The main risk is political contagion, i.e. that a perceived unjust outcome for Greece increases support for anti-EU parties in other countries. Spain and Portugal hold important elections later this year. However, Greek economic hardship may cause other electorates to become more risk averse. The euro remains popular across EMU countries, even where the EU is not.

• The Greek government has called a referendum for 5 July on EU / IMF proposals that have now been withdrawn. The government will campaign for “no” but nevertheless expected the EU / IMF to provide additional financial support this week. This was refused and the ECB has frozen emergency liquidity assistance (ELA) to Greek banks, resulting in their closure until the election.

• “Grexit” is still not the most likely scenario. Greeks will probably vote “yes” to signal their desire to retain the euro. The current government would probably then fall, resulting in a temporary “national emergency” coalition or new elections. Negotiations on a new deal would be difficult because the current Greek government has squandered any remaining goodwill and a magnified recession will inflict further damage on public finances and the banks. The ECB would keep ELA frozen until a deal was concluded, requiring capital controls to be maintained. The government would probably be unable to meet some spending commitments, issuing IOUs instead. An eventual deal would keep Greece in the euro and allow a slow lifting of capital controls, as occurred in Cyprus after its 2013 crisis.

• The initial reaction of bond markets to the Greek news has been smaller than feared, with 10-year Italian and Spanish spreads over Germany widening by about 30 bp. This partly reflects confidence that the ECB will modify and / or expand its QE programme to provide additional support for peripheral markets if necessary.

• The suggested scenario here, therefore, is that Greece will remain a source of negative news for the foreseeable future but the wider market implications will be limited because of a benign global liquidity / economic backdrop.

Chinese economic news is likely to improve over the summer but narrow money trends suggest that growth will remain sub-par.

The PBOC on Friday released additional monetary data for May, allowing the calculation of the “true” M1 measure monitored here. As previously explained, the official M1 series comprises currency in circulation and corporate demand deposits, i.e. it excludes household demand deposits. This is a major deficiency when policy is attempting to shift to consumption-led growth. True M1 corrects this omission.

The first chart shows six-month rates of change of industrial output and real (i.e. inflation-adjusted) true M1. The latter contracted in late 2014, correctly warning of recent economic weakness. Its six-month change has recovered in 2015 but current growth is no higher than a year ago. Six-month industrial output expansion may revive from about 2% currently to 3-4% during the second half (i.e. 6-8% annualised).

The second chart separates out corporate and household demand deposits. The latter have tended to lead in recent years and remain weak currently, suggesting that policy easing has yet to stimulate consumer spending. True M1 growth is being driven by corporate demand deposits, which include deposits of state-owned enterprises. A near-term economic pick-up based on higher corporate / state spending may not be sustained without consumer follow-through.

UK pay growth is picking up strongly, consistent with the historical relationship with the job openings or vacancy rate*, a measure of labour market tightness. This relationship suggests a continued upswing through mid-2016, at least.

The three-month smoothing of average weekly regular earnings (i.e. excluding bonuses) rose by 2.7% in the year to April, the largest annual increase since February 2009. Total earnings and earnings per hour also grew by an annual 2.7%. Hourly earnings had been lagging the weekly measure (i.e. employees’ average weekly hours had been rising) but the gap closed in April – see first chart.

A post in September argued that weekly regular earnings growth, then below 1%, was about to rise significantly, since the job openings rate had surged over the prior two years and had almost regained its pre-recession peak, indicating a tight labour market – see second chart, which is reproduced from the earlier post. The Bank of England and consensus view, by contrast, was that hidden slack (e.g. a large number of part-time employees wishing to work full-time) and rapid labour supply expansion would limit upward pressure on pay.

The weakness of earnings growth in 2013-14 seemed to support the consensus view that the job openings rate was overstating labour market tightness. The two series had turned down simultaneously in 2008 so their divergence in 2013-14 suggested a structural break in the relationship.

As noted in the earlier post, however, the openings rate led earnings growth by between six and 12 quarters at turning points in the 1970s and 1980s – third chart. The simultaneous fall in 2008, therefore, was unusual but the divergence between weak earnings growth and a surging openings rate in 2013-14 was not.

The current earnings pick-up is consistent with the historical relationship. As noted in the earlier post, “the openings rate bottomed in 2009 but embarked on a sustained rise only in the second quarter of 2012. Based on the average nine-quarter lag following the three increases over 1970-90, this suggests an upswing in earnings growth starting in the third quarter of 2014.” Regular earnings growth has risen steadily from a low of 0.7% in the second quarter of 2014 – fourth chart**.

The increase in pay growth is more impressive against a backdrop of falling / low consumer price inflation, which would be expected to subdue wage demands.

The job openings rate rose further in the fourth and first quarters, surpassing its 2008 peak, though has since fallen slightly. If the first quarter were confirmed to be a peak, the minimum six-quarter lead in the 1970s-80s would suggest a continued uptrend in earnings growth through the third quarter of 2016, at least.

*Vacancies as a percentage of employee jobs plus vacancies.
**Quarterly data except for the last data points, which refer to the latest three months. 

The balance of US activity news has improved so far in June, supporting the widely-held view that the economy is rebounding from weather- and strike-related weakness in early 2015. Monetary trends suggest that the pick-up will extend into late 2015 but investors should be wary of extrapolating optimism into 2016, for several reasons.

First, monetary strength may be fading. Six-month growth of real narrow money peaked in February, though appears to have remained solid in May* – see first chart. Real narrow money leads activity (industrial output) by nine months on average**, so this suggests that economic growth will moderate at end-2015.

Secondly, corporate finances are deteriorating. The “financing gap” of non-financial corporations, i.e. capital spending minus retained earnings, rose to 1.2% of GDP in the first quarter, the highest since the second quarter of 2008 – second chart. The gap is not yet at a worrying level but increases historically have preceded economic slowdowns or recessions, albeit sometimes with a long lead – third chart.

Note that the financing gap does not include borrowing to fund share buy-backs and M&A. A wider deficit measure including net equity buying tends to lead the yield spread between non-investment-grade corporate bonds and Treasuries and reached 4.3% of GDP in the first quarter, suggesting future spread widening – fourth chart.

Thirdly, cycle analysis warns of a possible economic downturn in 2016-17. This was discussed in detail in a post last August but, to summarise, the Kitchin stockbuilding cycle is scheduled to reach a low between 2015 and 2017, and the Juglar business investment cycle between 2016 and 2020. There is a risk that the downswings in the two cycles will coincide in 2016-17, producing a recession.

A recession should be signalled well in advance by a sharp slowdown or, more likely, a contraction in real narrow money – as previously discussed, such a contraction occurred before 10 out the 11 post-WW2 recessions, the exception being the 1954 downturn, which had a large fiscal element***. No such signal has yet been given.

*Based on May monetary data and a forecast of the consumer price index.
**Based on a study of G7 data over the last 50 years. The study – available on request – also confirms that real narrow money is superior as a leading indicator to real broad money, real bank lending and the rate of change of real bank lending (i.e. the “credit impulse”).
***Real government spending fell by 6.4% between 1953 and 1954.

US job openings (vacancies) surged in April, consistent with the forecast here of economic reacceleration from the spring, following a rebound in real narrow money growth in late 2014 / early 2015 – see previous post.

The job openings rate (i.e. vacancies expressed as a percentage of the total number of jobs) rose to 3.7% in April, the highest since 2001. The official openings series begins at end-2000 but earlier numbers can be estimated from help-wanted data compiled by former Fed economist Regis Barnichon. The current openings rate is close to the 2000 cycle peak of 4.0% – see first chart.

The job openings rate is inversely correlated with the unemployment rate, leading the latter at turning points by an average of four months since 1970. The level of openings forecasts the flow rate of unemployed people into employment. A rise in this flow could, in theory, be offset by an increase in layoffs. Weekly initial unemployment claims, however, remain low, as does the monthly layoff tally compiled by Challenger, Gray and Christmas.

Based on the average four-month lead, the further rise in job openings suggests that the unemployment rate will continue to fall until August, at least. Mr Barnichon’s model, indeed, predicts that the jobless rate will smash through 5.0% over the summer, reaching 4.5% by end-2015.

The job openings rate is positively correlated with pay growth, as measured by the annual change in the wages and salaries component of the employment cost index (ECI), with a six-month average lead at turning points since 1983, when the ECI series begins. The recent pick-up in pay growth, therefore, may extend into late 2015, at least – second chart.

A recovery in the Fed’s preferred inflation measure – the annual change in the personal consumption chain price index excluding food and energy – from its April level of 1.2% is probably also required, but labour market trends are consistent with a September interest rate rise.