Un mois de septembre chargé

Aux États-Unis, le rapport sur l'emploi du mois d'août a confirmé que le marché du travail se ralentissait, mais selon les normes historiques, il conserve toutes les caractéristiques d'un atterrissage en douceur : la création d'emplois reste positive et la croissance des salaires reste robuste. Néanmoins, cela a suffi à relancer la discussion sur le marché quant à la possibilité que la Fed recoure à une réduction de 50 points de base la semaine prochaine. Nous ne sommes pas convaincus. Commencer la phase d'assouplissement par une mesure aussi importante donnerait le ton à l'ensemble de la trajectoire - en termes de tarification du marché - alors que nous ne pensons pas que le FOMC ait pris une décision quant à la gravité du ralentissement à venir. Nous pensons qu'il serait plus simple de s'en tenir à une réduction de 25 points de base tout en indiquant clairement lors de la conférence de presse que la Fed n'hésiterait pas à procéder à une réduction « importante » et/ou « rapide » si le besoin s'en faisait sentir par la suite.

La BCE doit se décider avant la Fed. Bien qu'il y ait peu de suspense sur une réduction de 25 points de base ce jeudi, le marché se concentrera sur tout indice de « forward guidance » sur les prochaines étapes de la part de Christine Lagarde. Nous pensons qu'elle ne dévoilera pas ses cartes, étant donné que le débat au sein du Conseil des gouverneurs est toujours en cours. Les toutes dernières données jouent en faveur des colombes : les détails des comptes nationaux de la zone euro pour le deuxième trimestre confirment que les entreprises compensent de plus en plus la pression des coûts de la main-d'œuvre en réduisant leurs marges.

Avec une demande médiocre en Europe et un début de ralentissement aux États-Unis, les développements cycliques en Chine - en tant que « consommateur de dernier recours » potentiel pour l'économie mondiale - revêtent une importance particulière. Nous ajoutons à nos perspectives chinoises généralement prudentes une exploration d'un autre problème politique pour Pékin : la nécessité d'élargir l'assiette fiscale, maintenant que les ventes de terrains ne peuvent plus financer une grande partie des investissements publics, alors que les dépenses des ménages sont déjà faibles.

Enfin, la nomination d'un Premier ministre en France ne clarifie pas complètement la politique de Paris. Le projet de loi de finances pourrait être reporté de quelques semaines afin de permettre l'élaboration de compromis délicats.

Vous pouvez lire l'analyse de Gilles Moëc ci-dessous.

(Ever so) softly landing

In the US, the employment report for August has confirmed the message from July: the US labour market is now clearly softening, but without the markers of a “hard landing.”

Beyond the below expectation monthly print for the Establishment survey (142K versus 165K), the downward revision of the previous two months has made the pace of decline visibly steeper (see Exhibit 1). The market initially read into the August print a significantly higher risk that the Fed will have to contemplate “unusual action” to deal with the downturn, pricing up to 41 bps worth of cuts next week, against 33 bps before the release. Yet, expectations have been subsequently pared back, returning to 33 bps as of Friday evening. We agree with the latter and consider that prudence should be maintained on the quantum of easing the Fed is prepared to provide in one go at this stage.

True, some of the initial market reaction may have been magnified by the interview of Fed Governor Waller, who after the payroll release expressed his openness to “larger cuts”, and mentioned the possibility to cut fast, i.e. at every meeting, if need be. Yet, cutting by 50 bps is still a big step in our view, in a situation where available data still do not suggest a hard landing is underway. Indeed, we reiterate the point we made before the summer recess: the ongoing labour market adjustment remains very soft by historical standards. True, in August, the “Sahm point” was hit again – just: over the last 3 months, the unemployment rate has averaged 4.2% (4.13% in July), 0.57 points above the lowest point of the last 12 months (3.63% in the three months to August 2023). This would normally be a strong predictor of recession, but every time in recent history the “Sahm point” had been hit, job creation was much weaker than what is being currently observed (see Exhibit 2 – we took away the Covid period from the graph otherwise the scale of the shock would make any reading of the “ordinary recessions” very difficult). The labour market has turned a corner, that much is clear, but it is still a very soft adjustment. In addition, while job creation is slowing down, there is still quite a bit of acquired speed on wages. Indeed, hourly earnings rose by a stronger-than-expected 0.4% on the month, and by 3.8% on a 3-month annualised basis. ​

When poring over the details of Governor Waller’s interview, it is not obvious that his openness to “big moves” necessarily applies to the September meeting already. What makes us still favour a 25-bps cut next week, rather than 50 bps, is that opening the easing phase with a big step down would probably set the tone – at least in terms of market pricing – for the entirety of the trajectory, and such signal could be difficult to alter down the road. The level of confidence to do this would need to be high. In our view, such approach works only if one considers the FOMC has already made up its mind about the slope of the macroeconomic downturn, which we do not think is obvious given the recent dataflow. It would be easier, in our view, for the Fed to cut by “only” 25 bps next week while making it plain that the FOMC would not hesitate to raise the quantum of cuts at the next meetings and/or adopt a “quick fire” pace of cuts ​ if the dataflow indeed starts pointing to a harder landing. This approach would in our view encapsulate Waller’s points on cutting at “consecutive meetings” or “larger cuts” if it is appropriate, informed by the dataflow. But for now, we would borrow verbatim from Waller’s comments: “Based on the evidence I see, I do not believe the economy is in a recession or necessarily headed for one”.

Is there life after September?

While we expect the Fed to be quite clear on the generic “direction of travel” after its first hike, even if it will make it conditional on data, we are not confident we will get a lot of clarity on that front from the ECB upon delivering a second 25 bps cut on Thursday. While the cut itself makes little doubt – the forward contracts were pricing a probability of 99% to such outcome as of last Friday night – we would expect the market to focus on even the slightest traces of forward guidance in Christine Lagarde’s speech and Q&A. We explored last week how the hawks are maintaining a very prudent view of the macroeconomic developments in the Euro area, which would make it difficult for Christine Lagarde to be straightforward on the trajectory.

However, we think the very recent European dataflow has added more evidence that the ECB should sketch out a “decisive easing” trajectory. There was indeed a lot to mine in last week’s new estimate of Q2 GDP. The key point for us was not the downward revision from the first estimate (from 0.3% qoq to 0.2%), even it confirmed the Euro area is still not emerging from its spell of economic mediocrity. We were more interested in the details of the quarterly accounts on the wage and profit behaviour.

We already knew that negotiated wages had – at long last – finally slowed down in Q2, but last week confirmed that “actual” wages per head decelerated as well, to 4.3% yoy from 4.8% in Q1, hitting their slowest pace since the end of 2021 (see Exhibit 3). This may not look like much, but interestingly this was lower than in the ECB’s latest forecasts (5.1% in Q2). Now, nominal wages alone do not necessarily say much about the domestic inflation pipeline. It is the “wage, productivity and margin” nexus which matters – as Christine Lagarde made it plain during the July press conference. But we think on that front as well the news is getting better.

Taken at face value, it is very difficult to extract any signal from the quarter-on-quarter changes in unit labour costs (wage growth minus productivity gains) and unit profits, given their extreme volatility (see Exhibit 4). Still, when averaging them over four quarters, we think the picture gets quite clear: while unit labour costs have only started to slow down – and remain at a high pace – unit profits are now starting to contract. Businesses are now doing the opposite of the “greedflation” of 2022 and they are offsetting the labour cost push by lowering their margins. Incidentally, beyond the favourable impact this is having on inflation, the contraction in profits helps explain the deterioration in corporate investment which was confirmed in the new GDP estimate for Q2 – another reason for the ECB, in our view, to embark on a clear easing process.

Even if the state of the debate within the Governing Council suggests Christine Lagarde will not want, on Thursday, to elaborate too much on what could come next, her diagnostic on the current macro situation in the Euro area – informed by these latest data points – could err on the dovish side in our view. Another key input obviously will be the ECB’s new forecast batch. The ECB is likely to forecast marginally lower headline inflation in 2024's annual forecast to 2.4% (-0.1pt), but we see risks of an 0.1 percentage point upside revision to 2.8% for 2024 core inflation forecast. For 2025 and 2026, the ECB should account for lower momentum in oil prices (-6% in average), and stronger euro (+2%). This should result in important revision for headline inflation to 1.9% in 2025 (-0.3pp) and 1.7-1.8% (from 1.9%) in 2026, while core inflation revisions should be much more limited - expected at 2.1% (-0.1ppt) in 2025 and unchanged at 2% in 2026. Not a very strong signal, but the signal the ECB believes in a slightly faster convergence towards its target would still indicate a readiness to ease faster in 2025.

Taxing China

As we are readying for slower US demand in the short run (the magnitude of such deceleration is still debatable, but not the direction of travel there) it is natural to look for alternative sources of traction in the world economy. Europe “does not do demand”, so China would normally be the “natural candidate” to pick up the tab. Yet, habitual readers of Macrocast will be familiar with our reserved Chinese outlook. We explore here another side of the complex policy equation for Beijing: the collision of a structural need to expand the tax base with the necessity to reduce the saving ratio of households.

Indeed, what is striking in China is the low level of tax receipts as a percentage of GDP at less than 14% in 2022 according to the IMF. Tax ratios are normally a positive function of the degree of economic maturity of a given country, but even when compared with similar countries in terms of GDP per head or trade openness, China is close to the bottom of the distribution (the average in comparable countries stood at 19% (see link here https://www.elibrary.imf.org/view/journals/002/2024/050/article-A002-en.xml )and its tax ratio has been declining from a peak at 18% of GDP in 2011. Such thin tax base co-exists with a high public debt (around 100% of GDP).

One of the reasons taxes could be kept low despite increasingly active fiscal policies is that a significant share of local government revenue – at the forefront of the country’s public investment effort – comes from land-use sales (30% of their total income in 2022 according to the Peterson Institute – see link here). As long as China has been deploying an extensive growth model, such approach was seemingly virtuous. Local authorities would fund infrastructure spending by selling land at a rising price, and the very extension of the electrical grid, roads, sewage network etc… would mechanically raise present and expected land values. Moreover, local authorities do not forfeit the entirety of their property rights (legally, transfer of ownership of state-owned land is prohibited): they sell “land usage rights” to developers for a maximum of 70 years. The value of the asset side of the local authorities is not only boosted by the rising price of future land-use sales, but also by the long-term gains on the existing operations. In such a configuration, looking at gross debt could be misleading. The local governments’ liabilities should be compared with the rising value of their fixed assets. Looking at net debt is tempting … as long as demand for land grows.

In the “extensive” growth model, urban population rises as country dwellers flock to the cities, which can transitorily offset overall demographic dynamics which have been mediocre for a long time. But what happens once those reserves get exhausted, or when the general upgrade of the housing stock has been completed? Exploring land rent issues has been one of the first endeavours of economic science. The verdict is straightforward: pouring more capital (e.g. infrastructure) into land when demand is structurally falling amid declining population will simply result in diminishing, and soon negative returns: at some point, the cost of capital will exceed the value of the land. Land-related revenue of local government has been falling since 2022.

Our colleague Yingrui Wang has just released a comprehensive paper on the state of Chinese banking (see link here). One of her key findings is that the banks’ main problem in the current real estate correction may not be developers’ loans or mortgages, but rather their exposure to the special financing vehicles set up by local governments, which account for 13% of their balance sheet in 2022, against 3.8% for “direct” real estate operations. It is thus a crucial matter of general financial stability in China that the central government – or the central bank – support the local authorities.

Still, whatever “financial engineering” is put in place to mitigate the impact of the current real estate correction on financial stability in the short run, ultimately, what needs to happen is a proper overhaul of government income in China, substituting “ordinary taxation” to asset sales. A difficulty there is that “ordinary” government revenue would need to rise while households maintain high savings ratio to build up reserves for their retirement in the absence of a comprehensive public pension system. Currently, because of a large basic exemption, 70% of Chinese households do not pay any income tax, and although facially the top rate for income tax stands at 45% - a level seen in many mature countries – the average income tax rate is below 5.5%. Squeezing further an already anaemic consumption could be the net result of such overhaul.

Beijing could simply to postpone any large tax reform, and then simply accept to curtail public investment to adjust to the decline in land price. Either way, these issues make it very unlikely China will be in position to contribute massively to world demand. ​

France: delay into October?

The suspense is only partly over now that Michel Barnier has been appointed Prime Minister. Indeed, the “mother of political battles” remains the budget bill for 2025. Technically, it normally needs to be transmitted to parliament on 1^st October, but Pierre Moscovici, Chair of the French Audit Court and Public Finances High Council, in an interview over the weekend indicated that he could see a delay until mid-October. This may reflect some “backward counting” from the Constitutional rules. Indeed, the budget needs to be approved by December 31^st, and parliament has 70 days to examine it. Thus, October 20^th could be, from a practical point of view, the longest possible delay.

These few weeks of grace could indeed come handy since the political conditions around the budget remain delicate. Michel Barnier comes from the centre-right and his appointment could suggest that the business sector would be protected as much as possible from tax hikes which look increasingly unavoidable. Yet, he has also pledged to “put back on the table” some aspects of the latest pension reform, which has been a target of both the left and the Rassemblement National. Since the left has already announced they would stand in clear opposition to the new government – announcing they would table a motion of no confidence – the attitude of the Rassemblement National will be key. While they made it clear they would not “torpedo” the Barnier government immediately, they would probably call for a protection of public spending on strategic issues for them (security in particular), and they have already pledged to support consumers’ purchasing power in the budget, which would in principle make it impossible to cut social transfers or lift social contributions, VAT or income tax. Solving this political equation while still producing a budget bill consistent with the European surveillance framework – and the rating agencies’ expectations – will be extraordinarily delicate.