The strong rally in equities since late March contrasts with static longer-term government bond yields, causing some to argue that economic expectations in the two markets are out of sync, the suggestion being that a pessimistic bond market is smarter. This interpretation is not shared here: recovery expectations are modest in both markets, while a key common influence has been a fall in real discount rates driven by global “excess” money.

A static US 10-year Treasury yield conceals a recovery in inflation expectations offset by a fall in the real yield, which is testing its 2012 low – see first chart.

The inflationary component of the yield correlates with forward-looking activity measures, such as the ISM manufacturing new orders index – second chart. Treasuries, therefore, are discounting a modest economic recovery (modest because inflation expectations remain low by the standard of recent years).

Why has the real yield collapsed? The monetarist view is that real discount rates across the investment spectrum have been pushed down by global “excess” money and resulting increased demand for financial and real assets.

The third chart compares six-month changes in the 10-year real Treasury yield and G7 “excess” money growth, the latter measured as the gap between six-month rates of change of real narrow money and industrial output (plotted inverted to reflect the expected negative relationship).

The idea is that falls (rises) in real yields are driven by accelerations (decelerations) in excess money.

Note that the real yield measure used in the chart is the nominal 10-year yield minus the consensus 10-year CPI inflation forecast – this measure has a much longer history than the yield on inflation-protected securities (TIPS).

The chart supports the view that excess money is a key driver of real yields*.

Turning to equities, the strong rally reflects a rerating consistent with the fall in the real Treasury yield – fourth chart.

Analysts' earnings expectations are gloomy: the 12-month forward estimate has yet to bottom and implies a recovery of only 6.5% from the trailing level, which is down by 22% from end-2019 – fifth chart. So the consensus in equity markets, as with bonds, discounts a limited economic revival. Earnings estimates have substantial upside in the strong rebound scenario favoured here, involving global industrial output returning to its pre-crisis level by H1 2021.

The current favourable excess money backdrop for markets could become much less so during H2.

The sixth chart shows six-month rates of change of G7 real narrow money and industrial output, the latter projected forward assuming constant monthly growth sufficient to return output to 95% of its end-2019 level by December 2020. With real money growth unlikely to be sustained at its current extreme, the real money / output growth gap could turn negative during H2.

Recall that the direction of real yields is related to accelerations / decelerations of excess money. Excess money is certain to decelerate, whether or not its rate of change turns negative.

The suggestion is that real yields will back up during H2, in turn raising the possibility of a correction in equities unless earnings expectations rise significantly.

A counter-argument is that central banks will cap nominal yields, resulting in a further fall in real yields as inflation expectations continue to recover in line with improving economic prospects. They may want and try to, but can they? Japan's experience with yield curve control may be misleading – a downward trend in global yields gave the BoJ an easy ride. A vast expansion of QE might be needed to maintain current yield extremes, which might itself prove destabilising by unhinging inflation expectations.

*Statistical measures of the correlation are affected by several instances of extreme movements in excess money growth, including the present. One solution is to impose minimum and maximum levels of -4 and +4 percentage points respectively on the six-month excess money growth change. The correlation coefficient between the six-month real yield change and the adjusted excess money growth change was -0.49 over 1990-2019.

The global stockbuilding (inventory) cycle is judged here to have bottomed in H1 2020, probably Q1. The cycle acted as a drag on global economic momentum in 2018-19 but is now scheduled to provide a tailwind at least through end-2021.

Reasons for believing that Q1 marked the trough include:

1. A low was due – the cycle has averaged 3.5 years and last reached a trough in early 2016.
2. The negative contribution of stockbuilding to G7 annual GDP growth in Q1 was large enough to be consistent with a cycle low – see first chart.
3. A timely monthly indicator derived from G7 business surveys firmed in Q2, reaching a 13-month high in June – first chart.

A weakness of the analysis is that it excludes China, the largest industrial economy. Chinese industrial output has rebounded more strongly than retail sales, raising the possibility that stock levels have increased significantly, in turn suggesting that an inventory correction in China could offset a recovery in the G7 cycle.

China does not release quarterly national accounts inventories data but a business survey indicator can be constructed along the same lines as for the G7. This indicator has reached peaks and troughs around the same time as the G7 indicator, consistent with the stockbuilding cycle being global – second chart.

The Chinese indicator reached a low in February – further evidence that the global cycle troughed in Q1. It spiked higher in March / April, probably reflecting the production side of the economy reopening ahead of demand, but returned to negative territory in May / June.

The indicator, therefore, does not support claims of a significant inventory overhang. G7 and Chinese stockbuilding cycles appear aligned and should have reinforcing effects on global economic momentum in H2 2020 and 2021.

Global six-month real money growth – on both narrow and broad definitions – is estimated to have risen to another post-WW2 high in June, based on data for the US, China, Japan, Brazil and India, which have a combined two-thirds weighting in the G7 plus E7 aggregates calculated here.

A global leading indicator derived from the OECD’s country leading indicators usually mirrors monetary swings with a lag and posted a second large monthly rise in June, confirming an April low – second chart. The OECD indicators exclude money for most countries so represent an independent cross-check.

The equity analysts’ weekly revisions ratio, meanwhile, has normalised, consistent with the global manufacturing PMI new orders index rising further to 50-55 in July – third chart.

An April low in global economic momentum coincided with a low in the relative performance of “old economy” cyclical equity market sectors (i.e. materials, industrials, consumer discretionary, financials and real estate) versus defensive sectors (i.e. consumer staples, energy, health care and utilities) – fourth chart.

This old economy cyclical / defensive sector relative has exhibited a significant correlation historically with the US 10-year Treasury yield – fifth chart (correlation coefficient = 0.92 over 1995-2019). The recovery in the relative since April has opened up a wide divergence with a static yield, suggesting that one or other market is mispriced – monetary trends argue bonds.

The previous quarterly commentary suggested that the policy response to the covid-19 crisis would lead to a strong rise in global money growth, in turn suggesting strong economic growth in late 2020 / 2021. The monetary pick-up over the past three months has exceeded expectations, raising the prospect of a full-scale boom in 2021 with an accompanying inflation pick-up. “Excess” money may continue to buoy equities and other risk assets near term but could evaporate later in 2020 as a V-shaped economic rebound plays out.

The key monetary indicator used for economic forecasting here is the six-month change in real (i.e. inflation-adjusted) narrow money in the G7 economies and seven large emerging economies (the “E7”). This was at a modest level entering 2020 but surged over February-May, reaching the highest level on record in data extending back to the 1990s – see first chart. Allowing for an average nine-month lead, the suggestion is that global economic momentum will rise into early 2021, at least.

The global surge owes much to a 20.7% (not annualised) explosion in US narrow money since end-2019 but strong accelerations have occurred in almost all economies. China is lagging somewhat, reflecting more measured (and sensible) policy easing – second chart.

G7 plus E7 six-month real narrow money growth is probably at or close to a peak, which could signal a peak in economic momentum around end-Q1 2021. Base effects guarantee a slowdown in US nominal money by end-Q3, while G7 plus E7 six-month consumer price inflation is about to rebound in response to a recovery in the oil price. The expectation here, however, is that a further rise in Chinese money growth will take up some of the slack, maintaining global growth at a high level, with positive economic implications for later in 2021.

Narrow money is more useful for anticipating economic momentum turning points but broad money is a better guide to medium-term inflation prospects. Annual growth of G7 plus E7 nominal broad money, like that of narrow money, is the highest since the 1990s (at least) – third chart. The dual surge contrasts with late GFC developments, when a narrow money pick-up signalled an economic recovery but broad money weakness argued against an accompanying rise in inflation.

G7-only money data are available much further back. Annual broad money growth is the fastest since 1973 and not much below a post-WW2 peak in 1972 – fourth chart. US growth is above the corresponding peak and at its highest peacetime level since 1881.

The early 1970s experience is instructive. G7 broad money growth rose from a trough of 7.2% in Q1 1970 to 18.0% in Q4 1972. Consumer price inflation of 5.1% in Q4 1972 was lower than at the start of the money growth surge in 1970. The money growth / inflation gap had reached a record 12.9 percentage points and commentators were no doubt opining about a collapse in velocity. Inflation then embarked on a sustained surge to a mid-teens peak in Q4 1974, two years after the money growth peak.

The recent inflation fall, incidentally, is consistent with prior monetary trends – G7 annual broad money growth weakened during 2018, bottoming at 3.4% in November.

The inflation upswing of the early 1970s was magnified by a quadrupling of the oil price following the Arab oil embargo of October 1973. Absent a similar shock, a reasonable expectation is that G7 inflation will reach 6-7% in 2021-22. This assumes that current broad money growth – 16.9% in May – is partly absorbed by strong real GDP growth of 3-4%, while velocity falls by 7% pa – two standard deviations more than the average rate of decline over the past 50+ years.

Some commentary argues that the monetary surge reflects a precautionary drawing-down of credit lines by companies and will reverse as the economy normalises and loans are paid back, cutting short any increase in inflation. This argument is not supported by an analysis of the drivers of faster broad money growth. Monetary financing of fiscal deficits has contributed significantly more than increased bank lending – see table.

Much of the lending boost, moreover, reflects government schemes involving either future conversion to grants (the US Paycheck Protection Program) or guarantees on risky loans, a sizeable proportion of which is likely to be written off (e.g. the UK Bounce Back Loan Scheme). The grant / write-off element represents additional money-financed government spending, implying that the monetary financing numbers in the table are too low, and the lending numbers too high. Any negative impact on broad money from loan repayments, therefore, will be minor relative to the recent surge.

The consensus view that inflation will remain low or even weaken further assumes that the health crisis will result in a lasting negative impact on demand that outweighs damage to supply, implying a persistent negative output gap and labour market slack. The monetarist view, obviously, is that the demand impact, beyond the short term, will be positive not negative.

Recorded and hidden (by government schemes) unemployment has risen sharply but weakness, similarly, is likely to prove temporary. The US “permanent” unemployment rate – excluding those classified as being on temporary layoff – stood at 4.5% in June, below a long-term average (since 1985) of 5.1%. The suggestion of limited damage is supported by a still relatively favourable assessment of labour market conditions in the Conference Board consumer survey – fifth chart.

The global cycle analysis conducted here is consistent with the monetary signal of strong economic prospects and rising inflation risks. Recent business surveys confirm that the stockbuilding (inventory) and business investment cycles bottomed during H1, while an upswing in the longer-term housing cycle is regaining momentum in response to record low mortgage rates. Overlaying these activity cycles is the “long wave” price / inflation cycle, which averages 54 years and suggests a secular rise in inflation to a peak in the late 2020s.

The monetarist view that money is in excess supply helps to explain the strength of the bounce-back in equities and other risk assets in Q2. Historically, equities have outperformed cash significantly on average when global / G7 real narrow money growth has been above industrial output growth and / or above a long-run moving average. The real money / output growth gap was already positive while real money growth crossed above its long-run average in March.

Previous commentaries discussed a switching rule that holds equities only when both conditions are satisfied, defaulting to US dollar cash otherwise. The rule allows for data reporting lags and switches are made only at month-ends, so the latest move into equities was delayed until end-April. The sixth chart compares the hypothetical historical performance of the rule with that of being permanently invested in equities.

The rule is almost guaranteed to remain in equities during Q3 but a moderation of global six-month real narrow money growth coupled with a V-shaped economic rebound could result in the real money / industrial output growth gap turning negative around end-Q3, implying a Q4 switch into cash.

Any set-back in equities, however, is unlikely to mark the start of a bear market – these usually occur during stockbuilding cycle downswings but the cycle is now scheduled to be in an upswing until late 2021, at least. As discussed in previous commentaries, cyclical equity market sectors and emerging market equities usually outperform during such upswings, a trend that could be reinforced on this occasion by simultaneous accelerations in business / housing investment and associated strength in commodity prices.

May monetary numbers have now been released for Euroland, the UK and Canada, allowing calculation of G7 aggregates.

G7 annual broad money growth rose further to 16.9% last month, the fastest since 1973. The post-WW2 peak, reached in November 1972, was 18.3% – see first chart.

Annual narrow money growth had already surpassed its post-WW2 peak and rose further to 19.5%.

Money growth, broad and narrow, remains much higher in the US than elsewhere – second chart. Canada is in second place with Europe catching up rapidly. Interestingly, the UK has overtaken Euroland, having lagged since end-2016.

What level of future nominal GDP growth might be implied by current broad money expansion?

The monetarist rule of thumb is that monetary changes are reflected in demand and activity six to 12 months later and in prices after about two years. This suggests a peak impact on nominal GDP growth after about six quarters.

Between 1965 and 2018 the difference between G7 annual broad money growth and annual nominal GDP growth six quarters later averaged 1.4 percentage points (pp), i.e. a decline in velocity absorbed 1.4 pp of broad money growth.

Based on this average, current annual broad money growth of 16.9% might be expected to be associated with annual nominal GDP growth of 15-16% by late 2021.

Since current money growth is extreme, it is reasonable to allow for an extreme movement in velocity. The standard deviation of the difference between G7 annual broad money growth and annual nominal GDP growth six quarters later was 2.7 pp over 1965-2018. Assume a prospective difference two standard deviations above the average, i.e. 1.4 + 5.4 = 6.8 pp. Current money growth of 16.9% would then imply nominal GDP growth of about 10% in late 2021.

Assuming that G7 real GDP growth is limited to 3-4%, nominal GDP growth of about 10% would be consistent with inflation of 6-7%.

Note that these numbers relate to late 2021, by which time the positive base effect of the Q1 / Q2 2020 economic shutdowns will have dropped out of the annual GDP comparison, i.e. annual nominal GDP growth is likely to be 10%+ in Q2 2021 irrespective of any monetary boost.

The “quantity theory of wealth” described in previous posts implies an inverse relationship between broad money pass-through to nominal GDP and the wealth to income ratio, i.e. to the extent that the broad money surge is not reflected in faster nominal GDP growth, asset prices and wealth will rise by more over the long run.

The least likely outcome, according to this theory, is inflation weakness coupled with stagnation or deflation of asset prices – a scenario regarded as plausible by many investors.