Posts earlier in the year argued, based on narrow money trends, that the global economy would lose some momentum over the summer before reaccelerating in late 2017, led by the US. This forecast is on track, although the summer slowdown was milder than expected. Current monetary trends suggest another momentum peak in early 2018.

Changes in economic momentum are reflected in activity news surprise indices calculated by Citigroup and others. The Citigroup index covering the G10 developed economies peaked in January / March 2017 at its strongest level since 2011, turned negative between June and August but has surged since mid-September. The EM index has followed a similar profile but has been less volatile – first chart.

These movements were foreshadowed by monetary trends. The shift from positive to negative surprises over the summer followed a sharp fall in global six-month real narrow money growth between August 2016 and February 2017. Recent strong economic news, meanwhile, was signalled by a rebound in real money growth into June 2017 – second chart.

The surge in the G10 surprise index since mid-September has been driven by US economic data – third chart. US economic strength is consistent with a pick-up in real narrow money growth into August, discussed in several posts over the spring and summer. A post in June argued that the US surprise index would return to positive territory by the fourth quarter, converging with or overtaking the Euroland index – this remains likely.

June appears to have marked another peak in global real narrow money growth, which edged lower in July and August before falling sharply in September. The lead time between monetary trends and the surprise indices is variable but this suggests that economic news strength will fade in early 2018. Real money growth is around the middle of its range in recent years so is not yet signalling economic weakness.

Euroland coincident indicators remain strong but monetary trends suggest that economic growth will moderate in 2018.

The consensus interpretation of September monetary data, released yesterday, is likely to be upbeat. Annual growth rates of broad money M3 and narrow money M1 rose slightly, to 5.1% and 9.7% respectively, with the latter representing a 17-month high.

The forecasting approach here, however, focuses on non-financial monetary aggregates, i.e. excluding holdings of non-bank financial institutions – such holdings are less relevant for assessing near-term prospects for spending on goods and services. Headline M3 / M1 expansion has been supported in 2017 by a recovery in financial sector money growth following weakness in 2016. Annual growth of non-financial M3 has fallen from 6.1% in March to 5.0% in September, with non-financial M1 expansion retreating from 10.1% to 9.2% over the same period – see first chart.

The year-over-year figures, moreover, conceal a sharper decline in six-month growth rates. Non-financial M3 rose by 1.7% or 3.4 annualised in the six months to September, with non-financial M1 up 3.4% or 7.0% annualised; these increases are the smallest over six months since 2014 – second chart.

With six-month consumer price inflation recovering from a decline in the first half of 2017, six-month growth of the non-financial money measures expressed in real terms is also the lowest since 2014 – third chart.

Real money growth remains respectable by longer-term historical standards, suggesting a modest loss of economic momentum rather than weakness. Any slowdown, however, could surprise a consensus relying on continued loose ECB policy – confirmed by yesterday’s news of a nine-months-plus extension of QE – to sustain a strong economy.

Why are money trends cooling despite ECB support and gradual banking-system healing? One probable driver is rising bond yields – average 7-10 year government yields in July 2017 were 60 basis points higher than in July 2016, representing the largest year-on-year rise since 2011. The fall in yields in response to yesterday’s ECB news may prove short-lived against a backdrop of rising US yields and a weakening euro.

ECB research found that real non-financial M1 has outperformed other money / credit aggregates as a leading indicator of the economy, with real M1 holdings of non-financial corporations displaying a stronger correlation with future GDP expansion than those of households. Six-month growth of real corporate M1 deposits was above that of household deposits in September but has fallen sharply since June – fourth chart.

The country M1 deposit breakdown, meanwhile, hints at economic damage from Spanish political turmoil: six-month growth of real M1 deposits in September was the lowest since 2014. A rise in Italian growth, by contrast, suggests improving relative economic prospects – fifth chart.

Slowing QE may put downward pressure on the velocity of circulation of broad money in 2018, with negative implications for the economy and risk markets.

The 12-month rolling total of purchases of government and agency securities by the Fed, BoJ, ECB and Bank of England peaked at nearly $1.7 trillion in April 2017 and is projected to halve by mid-2018 – see first chart*.

Previous falls in the 12-month rolling total from peaks in December 2009, June 2011 and March 2014 were followed by significant declines in world equities starting 16, nine and 14 months later respectively – first chart.

Some monetarist economists claim that QE fluctuations have affected the economy and markets by changing the growth rate of broad money M3. The evidence does not support this claim. The initial burst of QE in 2009-10 was associated with a recovery in G7 M3 growth in 2010-11 but subsequent rounds appear to have had little impact – second chart.

QE, instead, may have mainly affected confidence / risk appetite and the velocity of circulation, rather than money growth directly. The annual rate of change of a monthly measure of broad money velocity, i.e. the product of industrial output and consumer prices divided by M3, rose after each of the three big QE surges – third chart.

The prospective QE slowdown, therefore, suggests renewed downward pressure on velocity in 2018. Unless broad money growth rises to compensate, this would imply lower nominal economic expansion and, probably, weaker markets.

Broad money trends are currently stable: G7 annual M3 growth is lower than a year ago but within its post-2012 range – second chart.

The forecasting approach here emphasises narrow rather than broad money. The demand to hold narrow money is closely related to spending intentions, so narrow money trends should reflect shifts in confidence / risk appetite and associated changes in broad money velocity. Global real narrow money growth has moderated recently, suggesting slower economic expansion in 2018 than 2017 – see previous post.

*The projection is based on Fed QE reversal plans, open-ended BoJ purchases of ¥5 trillion per month (i.e. close to the average over the 12 months to September), a final round of ECB QE of €25 billion per month over January-September 2018 and no purchases / sales by the Bank of England. Transactions in corporate securities are excluded.

The Chinese economy has cooled since early 2017 but growth remains respectable and current monetary trends are reassuring.

Nominal GDP is a better guide to economic fluctuations in China than official real GDP growth numbers, which often show little variation. Two-quarter growth of nominal GDP, seasonally adjusted, peaked in the first quarter of 2017, falling significantly in the second and third quarters. A slowdown had been signalled by a sharp drop in six-month narrow money expansion between August 2016 and February 2017 – see first chart*.

Narrow money growth, however, has stabilised since early 2017 at a solid level by historical standards. Current growth is much stronger than in late 2014 / early 2015, when the economy was in danger of entering a deflationary recession. Aggressive monetary and fiscal policy easing averted this risk and set the stage for the late 2016 / early 2017 boomlet.

Monetary pessimists highlight weakness in broad money M2 – six- and 12-month growth of this measure recently reached new lows in data extending back to the early 1990s. As previously discussed, however, M2 has been distorted by large swings in deposit holdings of non-bank financial institutions – movements in such deposits are unlikely to provide information about near-term prospects for spending on goods and services.

An additional suppressing factor since 2015 has been a switch by households and non-financial enterprises out of time deposits into money market funds and wealth management products, which have channelled the funds back to the banks via purchases of bank bonds. Such bonds are excluded from the M2 definition.

An adjusted broad money measure excluding financial deposits and adding bank bonds has been better correlated with nominal GDP than M2 in recent years – second chart. Six-month growth of this measure is below the historical average but has recovered since June and is higher than in late 2014 / early 2015.

Broad money, therefore, is giving a less upbeat message than narrow money but does not suggest economic weakness.

The expectation here, based on the above, is that nominal GDP growth will stabilise around its recent lower level through early 2018.

One risk to an optimistic view is that policy tightening during the first half of the year has yet to be fully reflected in monetary trends. Near-term economic resilience, moreover, may sustain recent upward pressure on inflation, causing further restrictive policy moves.

*True M1 = official M1 + household demand deposits.

Today’s news of a further rise in UK consumer price inflation to 3.0% in September supports the case for the MPC to raise interest rates in November but does not provide additional hawkish information relative to the August release.

CPI inflation averaged 2.82% for the third quarter as a whole versus a projection of 2.68% in the August Inflation Report. The MPC then expected a further rise to 2.75% in the fourth quarter; current trends suggest an average of 3.0%.

CPI inflation just made it to 3.0% in September: the unrounded figure was 2.96%.

The rise from August was driven by food inflation, which jumped from 2.3% to 3.4%. A post last month noted that CPI food inflation was unusually low relative to producer output price inflation of food products, which reached 6.0% in July and remained elevated at 5.5% last month. A further rise in CPI food inflation, therefore, appears likely.

The September data collection date preceded a 12.5% increase in the British Gas standard electricity tariff on 15 September – this will lift annual inflation by a further 5-10 basis points in the fourth quarter.

Core prices are still rising at an above-target pace. The core measure monitored here excludes energy, food, alcohol, tobacco and education, and adjusts for historical VAT changes. Seasonally adjusted, core prices rose by 0.2% between August and September and at a 2.6% annualised rate in the latest three months – see chart.

As previously discussed, the claim that the inflation overshoot is entirely attributable to sterling weakness has become more difficult to sustain following data revisions showing unit labour cost growth of 2.4% in the year to the second quarter.

The view here remains that the rising inflation trend stems from a significant pick-up in money growth from 2011, with this faster expansion subsequently contributing to the fall in the exchange rate – see previous post. Annual broad money growth appears to have reached a peak in late 2016 but has usually led inflation turning points by at least two years historically, suggesting that upward pressure will persist in 2018.

The IMF’s latest World Economic Outlook (WEO) – a useful summary guide to the views of the global policy-making consensus – is upbeat about economic prospects. Monetary trends, by contrast, suggest that GDP growth will slow in 2018. This is the reverse of the position a year ago, when the IMF consensus was gloomy* but money trends were signalling – correctly – that growth would surprise on the upside in 2017.

The IMF now expects its global GDP measure to grow by 3.6% in 2017, up from 3.5% in July. It forecasts a further increase to 3.7% in 2018, implying that growth next year will be the strongest since 2011.

The key forecasting indicator here is six-month growth of global real narrow money**. This appears to have fallen sharply in September, based on monetary data for countries accounting for two-thirds of the aggregate – see first chart. Six-month growth peaked most recently in June. Allowing for a typical nine-month lead, this suggests that the global economy will lose momentum around spring 2018.

The latest decline in six-month real narrow money growth follows a temporary slowdown in late 2016 / early 2017. Together, these have resulted a significant fall in year-over-year real money growth, from a peak of 9.8% in September 2016 to about 6% last month – second chart.

The monetary slowdown, as well as signalling a likely moderation of economic growth in 2018, implies a reduction in current liquidity support for markets. The gap between year-over-year real narrow money growth and industrial output expansion, monitored here as a gauge of the potential money flow into markets, has fallen to its lowest level since 2014 – second chart.

Another development suggesting reduced support for global markets is a sharp slowdown in Japanese portfolio investment overseas. On a 12-month rolling basis, Japan’s portfolio investment balance has swung from a deficit of ¥40.5 trillion or $370 billion in September 2016 to a small surplus in August 2017 – third chart. This mostly reflects a decline in Japanese purchases of foreign securities, which has coincided with a slowdown in QE since the Bank of Japan introduced its "yield curve control" policy – fourth chart

*The October 2016 WEO was entitled “Subdued demand: symptoms and remedies”.
**”Global” = G7 plus emerging E7. Real = adjusted for consumer price inflation. Narrow money = currency in circulation plus demand deposits or equivalent.