OECD leading indicator data for October released yesterday support the forecast here of a global economic “boomlet” in early 2017. Global real narrow money trends suggest that momentum will peak in the spring but are not yet warning of subsequent weakness.
The OECD presents its leading indicators in detrended form, meaning that a rise signals above-trend future growth. A post a fortnight ago argued that the G7 indicator would strengthen significantly in October, while monthly changes for July-September would be revised higher, based on data available then. The actual numbers were close to the estimates presented in the post – see first chart.
The indicators for all the G7 economies bar Italy increased in the latest three months, with the largest gains in the UK, Germany and the US. The indicators for China and other major emerging economies, meanwhile, started rising earlier and continued to strengthen in October – see the OECD's release.
The second chart shows six-month changes in G7 plus E7 industrial output, a trend-restored leading indicator derived from the OECD data and real narrow money. The recent pick-up in industrial activity was signalled by a rise in real money growth from August 2015. The positive monetary signal was confirmed by a rise in leading indicator momentum from December 2015.
Six-month real narrow money growth appears to have reached a peak in August 2016. Allowing for an average nine-month lead from money to activity, this suggests a peak in industrial output growth around May 2017. Such a scenario would be confirmed by a topping-out of leading indicator momentum in late 2016 / early 2017.
Real money growth has not fallen back sufficiently to suggest a significant economic slowdown after spring 2017. With leading indicator momentum still rising, a pro-cyclical investment stance probably remains warranted.
Simple “monetarist” rules for switching between global equities and cash continue to favour equities at present. A fall in G7 real narrow money growth, however, could trigger a “sell” signal during the first half of 2017.
Global equities have returned 6.6% in US dollar terms so far in 2016 (i.e. as of 6 December), or 7.6% assuming currency hedging. US dollar cash (i.e. three-month Eurodollar deposits) has returned just 0.8%.
Two “monetarist” rules suggested maintaining exposure to equities during 2016. The first rule holds equities if the annual growth rate of G7 real narrow money exceeds annual growth of industrial output. Global equities (unhedged) returned 10.0% per annum (pa) more than cash on average between 1970 and 2015 when the differential was positive*. They underperformed cash by 4.7% pa when real money growth was below industrial output growth. A negative differential, in other words, is associated with a significantly higher probability of equity underperformance, including outright weakness – see first chart.
One rationale for this finding is that faster growth of real money than economic output may indicate that a portion of current money holdings will be used to purchase financial securities, including equities, suggesting upward pressure on prices unless the supply of securities expands commensurately.
An alternative rationale is that faster real money growth indicates a future acceleration of output, with positive implications for equity earnings.
G7 real narrow money grew by 8.3% in the year to October while industrial output is estimated to have been unchanged, so this rule remains in favour of equities.
The second, simpler rule holds equities if G7 annual real narrow money growth is above its long-run average of 3% (after rounding**). Global equities returned 10.6% per annum (pa) more than cash on average between 1970 and 2015 when real money growth was above 3%. They underperformed cash by 6.2% pa when real money growth was below 3% – second chart.
Again, above-average real money growth may indicate either future purchases of equities or improving economic / earnings prospects.
With current G7 annual real narrow money growth above the 3% switching level, this rule also remains in favour of equities.
The third chart compares the performance of these switching rules with the return on equities relative to cash (i.e. buy and hold). Global equities outperformed US dollar cash by 3.1% pa on average over 1970-2015. The first switching rule, based on the real money / output growth differential, beat cash by 5.4% pa over this period, implying an excess return of 2.2 percentage points (pp, after rounding). The second rule outperformed by 6.0% pa, implying a 2.8 pp excess return.
The first rule was superior to the second rule during the first half of the period, capturing a greater proportion of equity market upside while offering similar protection against drawdowns. These characteristics changed during the second half of the period. The first rule, in particular, suffered a larger drawdown during the 2000-02 bear market***.
Both rules failed to switch into cash before the October 1987 equity market crash. The real money / output growth differential turned negative in October 1987 but the reporting lag implies that this would not have been known until December. Real money growth itself, meanwhile, fell below 3% only in December 1987.
Real money, however, was slowing sharply in spring / summer 1987, warning of rising risks and possible future “sell” signals. Annual growth fell from a peak of 9.9% in January to 5.9% by July – monetary data for July would have been available by end-August. This suggests supplementing the two rules with an additional defensive condition: switch into cash if G7 annual real narrow money growth falls by more than 3 pp within six months.
This modification “deals with” the 1987 “miss” and slightly improves the performance of the second rule at other times, with little impact on the first rule – fourth chart.
The additional defensive condition could conceivably trigger a “sell” signal in early 2017. G7 annual real narrow money growth has yet to decline – October’s reading of 8.3% was the strongest since August 2015. Six-month growth, however, peaked in August, with the monthly increase falling to only 0.2% in September and October – see fifth chart. The recent slowdown partly reflects faster consumer price inflation, which may be sustained by rising commodity prices.
Assume, for illustration, that real narrow money continues to increase by only 0.2% per month. Annual growth would fall to 4.7% by April, meeting the condition of a greater-than-3 pp decline within six months. Allowing for a one-month reporting lag, the rules would switch into cash at end-May.
This, to emphasise, is an illustration, not a forecast. Monthly narrow money growth could rebound in late 2016 / early 2017. Current data do not suggest an imminent risk to equities from a changing monetary environment.
*All performance figures quoted allow for reporting lags but are based on revised rather than original monetary / economic data.
**G7 real narrow money rose by 3.3% pa on average over 1970-2015.
***Both rules switched back into equities prematurely. The period was unusual because valuations were still high when the monetary signals turned positive. This suggests modifying the rules to include a maximum valuation condition for a “buy” signal.
Eurozone broad money M3 growth has been depressed by a record outflow of capital from the region over the past 12 months. Monetary trends, nevertheless, are consistent with respectable economic growth.
Annual M3 growth fell from 5.1% in September to 4.4% in October, the lowest since March 2015. Accelerating credit to the private sector and a rising government contribution, reflecting the ECB’s QE programme, have been pushing up on M3 growth recently but have been more than offset by a fall in banks’ net external assets – see first chart.
The change in banks’ net external assets is the counterpart of the balance of payments current and non-bank capital account. The net asset position declined by €351 billion in the 12 months to October despite a €337 billion current account surplus in the 12 months to September (the latest available month). This suggests a net outflow on the non-bank capital account of about €690 billion over the past year, equivalent to 6.5% of Eurozone annual GDP.
What explains the capital haemorrhage? It is tempting to blame fiscal / structural policy paralysis, banking sector woes and political uncertainty. QE, however, is likely to have been the main driver. Evidence from other countries that have conducted QE suggests that a large portion of liquidity created has flowed abroad. The domestic monetary impact has been small or even negligible, although QE may have contributed to stronger global broad money growth.
Does the M3 slowdown signal deteriorating economic prospects? Probably not. The forecasting approach here focuses on non-financial money, i.e. held by households and non-financial corporations. The headline M3 and M1 measures have been depressed by a fall in financial sector deposits, probably reflecting weaker demand for money due to low / negative rates, with no implication for spending in the economy. Annual growth of non-financial M3 was significantly stronger than that of M3 in October, at 5.2%, down from 5.7% in September. Annual non-financial M1 growth of 8.7% similarly exceeded headline M1 expansion of 7.9% – second chart.
Six-month growth rates of real non-financial M1 / M3, admittedly, have declined notably since early 2016 – third chart. Both measures, however, are around the middle of their ranges since 2013, a period during which GDP growth has averaged about 1.5% annualised – above most estimates of potential. Economic performance in late 2016 / early 2017 may be boosted by a rise in net exports, reflecting global strength and a competitive exchange rate.
UK money trends remain solid, arguing against the consensus forecast of an economic slowdown in 2017. Narrow money holdings of both households and private non-financial corporations (PNFCs) are expanding robustly, suggesting that consumers plan to increase spending despite talk of a coming real wage squeeze, while businesses are not yet reining in expansion in response to Brexit uncertainty.
The preferred narrow and broad money measures here are non-financial M1 / M4, covering household and PNFC holdings but excluding volatile financial sector deposits, which are less relevant for assessing near-term spending prospects. Non-financial M1 rose by 0.7% in October, while non-financial M4 increased by 0.3%. Annual growth rates were 9.9% and 6.4% respectively, down from 10.1% and 6.8% in September but high by post-crisis standards.
Six-month growth rates of real (i.e. consumer price-adjusted) non-financial M1 / M4 have risen since mid 2015, with this pick-up subsequently reflected in faster economic expansion. The real money growth measures are below recent peaks but remain solid, suggesting that GDP / gross value added (GVA) will continue to rise at an annualised 2.0-2.5% pace – see first chart.
The sectoral breakdown is also reassuring. Six-month growth of real household M1 picked up strongly in early 2016, correctly signalling recent consumption buoyancy. It has moderated recently but remains consistent with respectable spending expansion. Corporate real M1 growth, by contrast, has rebounded from weakness in early 2016, suggesting that firms are revising up investment plans in response to stronger-than-expected economic expansion and favourable financing conditions, despite Brexit uncertainty – second chart.
Annual growth of UK nominal (current-price) GDP increased further to 4.0% in the third quarter, according to preliminary ONS data, with monetary trends suggesting a further pick-up into early 2017, at least. The latest OBR Economic and fiscal outlook, for comparison, forecasts nominal GDP growth of only 2.8% in 2017.
Nominal GDP growth has risen since the third quarter of 2015, following increases in annual rates of change of narrow and broad money – as measured by non-financial M1 / M4 – from June and February 2015 respectively. Growth of both money measures continued to climb into September, with broad money expansion the fastest since 2008 – see chart. (October monetary data will be released on 29 November.)
Quarterly growth of constant-price GDP was left unrevised at 0.5% in the third quarter, against a suggestion here of an upgrade – see previous post. As expected on the basis of turnover data, services output rose further in September but the increase was moderated by a fall in financial services activities (not covered by the turnover survey), while July output was revised lower. The implied level of GDP in September, however, is 0.2% above the third-quarter average, creating a solid base for current-quarter growth.
The expenditure details show a 0.9% increase in business investment last quarter, consistent with resilient corporate narrow money trends. Consumer spending continued to drive overall expansion, rising by 0.7%. The ONS headlines refer to net trade adding 0.7 percentage points to quarterly growth but this is misleading: the improvement was entirely due to the balance of trade in non-monetary gold swinging from a deficit in the second quarter to a surplus in the third quarter, i.e. UK residents were net buyers of gold in the second quarter but sold in the third, perhaps partly reflecting profit-taking following sterling weakness. Adjusting for transactions in gold and other valuables, domestic demand contributed to 0.7 pp to third-quarter growth, with net trade subtracting 0.2 pp.