The assessment here remains that the global economy has entered a “double dip” currently focused on manufacturing but likely to extend to services / labour markets, reigniting worries about a hard landing. Economic weakness is expected to be accompanied by an inflation undershoot into H1 2025.

DM flash manufacturing PMI results for August were mixed across countries but on balance weak, suggesting a further small reduction in global manufacturing PMI new orders following a July plunge to below 50 (assuming no change for China and other non-flash countries) – see chart 1.

Chart 1

Services results were again much stronger than for manufacturing but there are hints of emerging weakness in a fall in output expectations since May and a drop in US / Eurozone employment indices to below 50 this month.

A previous post suggested that the OECD’s US composite leading indicator has reversed lower since publication of the last official data point, for June. An update based on partial data points to a further decline in August – chart 2. The OECD will release July / August data for its indicators on 5 September.

Chart 2

The OECD’s Chinese leading indicator has been falling since late 2023 and the decline is estimated to have continued in July / August – chart 3.

Chart 3

Weaker economic momentum and pricing power are feeding through to company earnings. Revisions ratios have turned down since April in the US, Eurozone and UK, with the August Eurozone reading the weakest since 2020 – chart 4.

Chart 4

By MSCI World sector, August revisions ratios were most negative in consumer discretionary followed by energy, consumer staples and materials. The ratios for consumer discretionary and staples were the weakest since 2020, suggesting that a fall-off in consumer demand has been a key driver of the renewed downturn in manufacturing – chart 5.

Chart 5

This week’s announcement by the BLS of a preliminary 818,000 or 0.5% downward revision to the March 2024 level of non-farm payrolls, meanwhile, raises the possibility that US employment has already stalled.

The revision is based on the comprehensive Quarterly Census of Employment and Wages (QCEW). A monthly QCEW employment series is available through March but is not seasonally adjusted. Chart 6 compares the monthly change in non-farm payrolls, as currently reported before incorporating the revision, with the change in a seasonally-adjusted version of the QCEW measure.

Chart 6

The increase in non-farm payrolls was 133k per month higher than growth of the seasonally-adjusted QCEW series during Q1. If overstatement of this magnitude has continued since Q1, reported growth of 108k and 114k in non-farm payrolls in April and July could imply small declines in “true” employment in those months.

The most important issue in the global economic outlook is the meaning of Chinese monetary weakness.

Six-month rates of change of narrow / broad money, bank lending and total social financing (on both new and old definitions*) reached record lows in June / July – see chart 1.

Chart 1

Monetary weakness has been entirely focused on the corporate sector: M2 deposits of non-financial enterprises plunged 6.6% (13.6% annualised) in the six months to July (own seasonal adjustment) – chart 2.

Chart 2

Recent regulatory changes appear to account for only a small portion of the corporate broad money decline.

A clampdown on banks paying interest above regulatory ceilings has resulted in a shift out of demand deposits but money has largely stayed in the banking system – available data suggest modest inflows to wealth management products and other non-monetary assets.

The clampdown has also discouraged the practice of “fund idling” (round-tripping in UK monetary parlance), whereby banks offered loans to corporate borrowers to meet official lending targets, with borrowers incentivised to hold the funds on deposit.

If an unwinding of such activity accounted for the decline in corporate money, however, short-term bank lending to corporations would be expected to show equivalent weakness. Such lending has continued to grow, albeit at a slower pace recently, as have longer-term loans.

A trend decline in the ratio of corporate M2 deposits to bank borrowing, therefore, has accelerated – chart 3.

Chart 3

Household money holdings, by contrast, have been growing solidly – chart 2. An alternative explanation for the corporate money decline is simply that households are still hunkering down as the property crisis deepens, with weakening demand for consumer goods / services and housing transferring income and liquidity from the corporate sector.

The latest PBoC and NBS consumer surveys confirm rock-bottom sentiment – chart 4. If this explanation is correct, corporate money weakness may presage a collapse in profits – chart 5.

Chart 4

Chart 5

Why hasn’t the PBoC hit the panic button? Policy easing has been constrained by currency weakness: the most comprehensive measure of f/x intervention (h/t Brad Setser) reached $58 billion in July, the highest since 2016 – chart 6. The recent yen rally has offered some relief, reflected in a narrower offshore forward discount, but the authorities may be concerned that this will prove temporary.

Chart 6

The strange policy of trying to push longer-term yields higher against a recessionary / deflationary backdrop may represent an attempt to support the currency, rather than being motivated primarily by concern about financial risks. To the extent that the policy results in banks selling bonds, however, the result will be to exacerbate monetary weakness and economic woes.

*The previous definition excludes government bonds so is a measure of credit expansion to the “real economy”.

A sharp fall in the global manufacturing PMI new orders index in July confirms renewed industrial weakness. The companion services survey, however, reported an uptick in the new business component, which is close to its post-GFC average. Will services resilience sustain respectable overall growth?

The understanding here is that economic fluctuations originate in the goods sector, reflecting cycles in three components of investment – stockbuilding, business fixed capex and housing. Multiplier effects transmit these fluctuations to the services sector – there is no independent services cycle.

The manufacturing new orders and services new business indices have been strongly correlated historically, with Granger-causality tests indicating that the former leads the latter but not vice versa*.

Several considerations suggest that the recent divergence will be resolved by the services new business index moving lower:

1. The services future output index correlates with new business and fell to an eight-month low in July – see chart 1.

Chart 1

2. Recent new business readings have been inflated by strength in financial services – chart 2. Financial services new business correlates with stock market movements, suggesting weakness ahead.

Chart 2

3. Consumer services new business correlates with the manufacturing consumer goods new orders index, which fell below 50 in July – chart 3.

Chart 3

Output price indices for consumer goods and services support the optimism here about inflation prospects through mid-2025. A weighted average has fallen back to its October 2009-December 2019 average, a period in which G7 annual CPI inflation excluding food / energy averaged 1.5% – chart 4.

Chart 4

*Contemporaneous correlation coefficient since 1998 = +0.84. Granger-causality tests included six lags. Manufacturing terms were significant in the services equation but not vice versa.

A post in June suggested that a recovery in the OECD’s US composite leading indicator was ending. A calculation based on the latest input data confirms a reversal lower.

The historical performance of the OECD indicator compares favourably with the Conference Board leading index. The OECD indicator recovered from early 2023, signalling that recession risk was (temporarily?) receding, while the Conference Board measure continued to weaken.

The latest published data point, for June, was released in early July. The next update is due on 5 September and will provide July / August numbers.

Chart 1 shows the published series (black), a replica series calculated here based on data available in early July (blue) and an updated replica incorporating an additional month of input data (gold). The updated series has fallen sharply from an April peak.

Chart 1

The decline reflects weakness in four components: consumer sentiment, durable goods orders, the manufacturing PMI and housing starts. The two financial components – stock prices and the 10-year Treasury yield / fed funds rate spread – were still marginally positive in July but levels so far in August imply a turn lower.

The price relative of MSCI World cyclical sectors, excluding tech, versus defensive sectors has mirrored movements in the OECD US leading indicator historically – chart 2. Relative valuation is high versus history and has diverged from a weakening global manufacturing PMI – chart 3.

Chart 2

Chart 3

Manufacturing PMI results for July support the forecast of a global “double dip” into early 2025.

The global manufacturing PMI new orders index plunged by 1.9 points from June to 48.8, a seven-month low. The combination of a one-month fall of this magnitude or greater and a sub-50 reading occurred in only 14 months since 1998, highlighted by shading in chart 1.

Chart 1

In chronological order, those months were:

• October 1998 (Asian / Russian / LTCM crises)

• December 2000 / January 2001 (start of US / global recession)

• September / October 2001 (911 terrorist attack)

• March 2003 (Iraq invasion)

• September through December 2008 (GFC climax)

• November 2011 (Eurozone crisis / recession)

• February through April 2020 (covid recession)

So the current signal suggests significant economic weakness and risk-off markets, at least until policy-makers respond.

The forecast that global economic momentum would weaken in H2 2024 was based on a fall in six-month real narrow money momentum into a low in September 2023 and an observation that the money-activity lag has recently extended to a year or more – chart 2.

Chart 2

The September 2023 real money momentum low suggests that PMI new orders will reach a low by January 2025. With money trends still weak, however, a recovery may be lacklustre.

Could PMI new orders break below the low of 46.5 reached in December 2022? The low in six-month real narrow momentum in September 2023 was beneath the preceding low in July 2022 – chart 2. Current weakness is more likely to spill over into labour markets, creating negative feedback loops.

“Surprise” economic deterioration is forecast to be accompanied by sharply weaker inflationary pressures, reflecting broad money stagnation in H2 2022 / H1 2023. The consumer goods PMI output price index fell back below its pre-pandemic average in July, following a plunge in the consumer services index the prior month – chart 3.

Chart 3

Will the Bank of Japan’s latest attempt to exit ZIRP prove any more successful than its previous two efforts, in 2000 and 2006?

The monetary backdrop is no more promising. The six-month rate of change of broad money M3 was 0.5% annualised in June compared with 1.3% and -1.1% respectively before the August 2000 and July 2006 rate hikes – see chart 1.

Chart 1

Money growth, admittedly, has been depressed by recent record intervention to support the yen. The judgement here is that the authorities have marked out a major low in the currency – see previous post – so f/x sales are likely to slow / end.

A reduced intervention drag, however, will be offset by a contractionary monetary influence from QT. The announced phased reduction in monthly purchases implies that the BoJ’s JGB holdings will fall by about ¥8 trn during H2 2024, equivalent to an annualised 1.0% of M3.

Credit developments are superficially more supportive of policy tightening. The six-month rate of change of commercial banks’ loans and leases was 3.5% annualised in June compared with -1.9% and 2.8% before the 2000 / 2006 hikes.

Bank lending, however, is usually a lagging indicator of economic momentum, suggesting a slowdown ahead in response to recent activity weakness.

The BoJ “will … continue to raise the policy interest rate” if its outlook for economic activity and prices is realised. Headline and core CPI inflation are projected to be close to the 2% target in fiscal years 2025 and 2026 based on the output gap turning positive and a “virtuous cycle between prices and wages continuing to intensify”.

The “monetarist”  forecast, by contrast, is that inflation is heading for a big undershoot. Six-month core CPI momentum was 1.5% annualised in June*, with lagged broad money growth suggesting a further decline into 2025 – chart 2.

Chart 2

Coming Japanese inflation experience will be another test of forecasting approaches. Simplistic monetarism has trounced new Keynesian orthodoxy so far this decade. Another win for monetarist simpletons will spell third time unlucky for the BoJ.

*Own estimate adjusting for policy effects and seasonals.