Accélérer

# 228 – 10 juin 2024

Les partis historiques de gouvernement conservent leur majorité au Parlement européen, malgré les gains de l'extrême droite, ce qui, compte tenu de leur longue tradition de coopération, devrait empêcher tout changement radical de politique à Bruxelles. Cependant, la décision du président français de convoquer des élections nationales anticipées est une source d'incertitude. Nous pensons que cette décision précipite l'affrontement entre les partis traditionnels et le Rassemblement national qui, de toute façon, allait encadrer les prochaines élections présidentielles en 2027.

L'agenda politique européen a été en grande partie suspendu pendant la campagne. Cette semaine, nous nous concentrons sur l'enquête antisubventions de la Commission européenne sur les importations de véhicules électriques en provenance de Chine. Bien que cela semble aligner l'UE sur l'approche combative des États-Unis, nous pensons qu'il y a ici les germes d'une « voie médiane » envers Pékin, qui est alimentée par des préoccupations sur les coûts d'une guerre commerciale - nous profitons de l'occasion pour souligner les estimations par l'Institut Peterson du coût pour les États-Unis des hausses de tarifs préconisées par Donald Trump - ainsi que tout simplement par les difficultés au sein de l'UE à former un front uni sur cette question. Nous constatons que même dans le camp populiste, les contradictions abondent en matière de protectionnisme en Europe.

Nous analysons la conférence de presse de la BCE de la semaine dernière. Comme nous nous y attendions, Christine Lagarde s'est montrée très prudente quant aux prochaines étapes après avoir réduit les taux de 25 points de base. Le marché ne prévoit plus trois baisses de taux complètes cette année. Il s'agit toujours de notre référence. Cependant, le fait que la BCE dépende des données peut être une source de volatilité, d'autant plus que la Fed est également confrontée à un flux de données particulièrement déroutant, illustré une fois de plus par un rapport contradictoire sur les salaires. Le marché se concentrera sur le nouveau « dot plot » du FOMC cette semaine. Nous nous attendons à ce que la « prévision médiane » passe à deux réductions pour cette année, contre trois en mars. Il existe un risque qu'il n'y en ait plus qu'une, mais nous pensons que cela supprimerait beaucoup trop d'options, car cela enverrait le message que la Fed ne pourrait pas procéder à une réduction avant les élections.

Vous trouverez ci-dessous l'analyse de Gilles Moëc.

Hastening the showdown

The European elections result were not expected to trigger any major shift of stance in the EU. Even if the overall political space of the mainstream parties is shrinking, the European parliament has a long tradition of cooperation between the main groups (centre-right, liberals and social democrats) which, based on the national results seen so far, are together likely to retain a comfortable majority of the seats. It may become more difficult to cut the necessary compromises and the changes in public opinion will have to be taken on board, but no radical change of direction could be expected. Still, of course the decision by the French President in response to the European vote to dissolve the National Assembly – the lower house of French parliament – changes the perspective.

Our first reaction is that this decision is hastening a “showdown” which was in any case shaping up for the next presidential elections in 2027, since the candidate of the far-right Rassemblement National was highly likely to be – just like in the last two occasions – one of the two contenders in the second round, with – according to the polls – a significant chance to win. A crucial issue for the current French government is that, deprived of an absolute majority in parliament since 2022, legislating has become increasingly difficult, often left hostage to varying “circumstantial majorities”. The usual tool at the disposal of the government – forcing the adoption of legislation without a vote unless defeated by a vote of no confidence – is powerful but still left it at risk. 5 years of complex parliamentary manoeuvring could ultimately weaken the position of the entirety of the mainstream parties further and build up even more support for the populist propositions. By triggering a snap election on 30 June and 7 July, President Macron is probably seeking a clarification, and aims at rebuilding a more solid mainstream majority by forcing the electorate into a binary vote for or against RN.

It is still a risky choice of course. The various families of the far-right have totalled around 38% of the votes in the European elections, while the historical government parties gathered only about 36.5%. ​ It would however be wrong to directly project the results of the European elections – a one-round, proportional vote – to the two-round majoritarian system of the parliamentary elections for which the immediate stakes are more obvious to most voters. The usual limitation of the RN constituency candidates in these elections is their difficulty to tap into any “reserve of votes” in the second round.

The situation has got more complicated since 2022 in the sense that the electoral fortunes of the French centre-left have improved notably – at least judging by the results of the European elections. The “mainstream vote” in the first round could this time be split three-ways (centre-left, Macron’s party and the centre-right) while last time it was split two ways, creating potentially more difficulties for mainstream voters to converge around a single candidate in each constituency in the second round. Traditionally, the French centre-left party (Parti Socialiste) enters an electoral pact with parties further left. It may be difficult for more right-leaning mainstream voters to support candidates representing “the whole left” in the second round in the constituencies where the centre-right will be eliminated in the first round (with symmetrical issues for left-leaning voters in constituencies where the centrists or the centre-right candidates will oppose RN in the second round). In any case, even if mainstream parties were to win a majority of the seats in July, building a stable majority may not be easier than today given the divergences between the centrists, the centre-left and the centre-right.

No clarification would ensue, and we would have to brace for similar moment in 2027 again. At the same time, we would note that, even if RN were to win a majority in the National Assembly, this would usher in a phase of “cohabitation” with the centrist President, which, given the strong powers granted by the Constitution, would probably reduce the capacity of a RN Prime Minister to trigger a massive policy shift, especially in the realm of defence, foreign and European affairs.

International trade issues to come back to the fore

Beyond the unexpected consequences for French politics, last weekend elections also mark the re-starting of the European legislative agenda – even if we strongly suspect that everyone in Brussels will wait for the result of the French national elections before making any big decisions. While we expect the commentariat to get fully absorbed by the “personnel issues” (the composition of the new European Commission), in the macroeconomic realm we think a key event to watch will be the publication by the Commission of the conclusions of its anti-subsidy investigation into imports of battery-powered cars from China launched in October 2023. We were struck, during the campaign, by how a general questioning of free-trade principles has generalised across all political families in the EU well beyond the populist groups. The terms of the debate in Europe remains however very different from the US. Over there, “blanket protectionism” (Trump) is competing with “targeted measures” (Biden). In Europe, even targeted action is not necessarily consensual, since some countries maintain a quite strict adherence to free trade principles, despite the push from populists.

Let’s start with reviewing again US-style “blanket protectionism. The Peterson Institute has recently proceeded to estimate the macro impact of the generic 10% customs tariff advocated by Trump (see link here). In the first approximation, assuming no margin accommodation – and older experiments with targeted hikes suggest it is unfortunately plausible – the cost to the economy is equal to the share in GDP of the imports subjected to the change in tariff multiplied by the change in tariff. The 10% generic tariff would thus cost 1% of GDP. When the authors from the Peterson Institute add the 60% “special treatment” for Chinese products, the cost rises to 1.8% of GDP. They note that not all of this would end up in the treasury coffers, thus making it unlikely higher tariffs could cover the fiscal cost of prolonging the tax cuts granted in 2017. Indeed, the rise in the price of foreign goods on American soil would provide cover to US-based producers to raise their market share.

This would however only have a limited offsetting impact on the purchasing power of US consumers, since these producers would either come out with higher production costs than the foreign incumbents or have a strong incentive to raise their profit margins. This explains why a customs duty triggers income transfers: from consumers to the producers and the government. This is not a zero-sum game however, since there is in any case a loss of economic efficiency: some domestic producers can capture a rent, equivalent to the difference between the new, higher price at the consumer level, and their production costs. This rent is only imperfectly re-injected in the economy (a significant share will be saved). This is one of the channels through which protectionism is socially regressive, even more so if it is supposed to “fund” the prolongation of tax cuts which mostly benefitted those at the upper end of the income scale. Besides, faced with higher costs, consumers will likely demand a rise in their nominal wage which will weigh on the profitability of the entirety of the domestic producers, including those who did not benefit from the protectionist tariff (e.g., those operating in the services sector). Protectionism results in inefficient resource allocation.

Besides, nothing would in principle prevent foreign producers from benefitting from a depreciation in the exchange rate of their currency which would offset the impact of the tariff. Eichengreen and Irwin in a well-known paper from 2009 (see link here) connected exchange rate policies and protectionism. Looking at the interwar disruptions, they noted that the countries which fought tooth and nail to preserve their anchoring to the Gold Standard ended up imposing more trade tariffs than those who “gave up” and let their currencies float. This result is quite intuitive: the “Gold Standard faithfuls” had to face a constant erosion in their competitive position as their currency was appreciating and tried to mitigate this by resorting to trade barriers. Paradoxically, those who advocated macroeconomic orthodoxy to repair the world economy after the disaster of the first world war ended up destroying what had been one of the greatest achievements of the 19th century: the advent of a “first globalisation” thanks to free trade. The US permanently gave up on the Gold Standard in 1971- which killed it for good - but it is highly likely that it is precisely the general conversion to floating exchange rates in the 1970s which sowed the seeds of the second globalisation. It is indeed quite miraculous that the world economy, despite the dramatic crisis triggered by the two oil shocks, ultimately avoided the contraction in global trade which had characterised the 1930s.

The interconnection between the FX regime and protectionism is becoming relevant again. The US routinely accuse its trading partners of artificially keeping their currency weak, or at least to hamper the normal market mechanisms governing exchange rate movements. It is however clear to us that in the current configuration, the strength of the dollar is largely the product the US economy outperformance relative to the other Western economies reflected in a generally more hawkish stance for monetary policy. But precisely, the implementation of Trump’s agenda would probably push the currency even higher, since the Fed would have to deal with a net fiscal stimulus compounded by another surge in consumer prices triggered by the trade tariff. Symmetrically, the countries at the receiving end of the US tariff hike would face even softer growth, favouring more monetary policy easing. The risk there is that a vicious circle materialises. As at least some of the effect of the tariffs on imports would be offset by the deterioration in US competitiveness brought about by the dollar appreciation, there would be a temptation in the US to “double down” on the tariffs, and so on and so forth.

How would the EU fare in these circumstances? We have already mentioned in Macrocast how a de facto closure of the US market to Chinese producers in case of a 60% levy would incentivise them to look to Europe as an even more crucial market, flooding the EU with products which most national European governments consider as strategic. This would be even more damaging if the Chinese currency were to depreciate not just relative to the US dollar but also to the euro, which would make sense since Chinese products would face a higher tariff than European ones. Then the protectionist tendencies in Europe would be spurred further, and the spiral we described for the US could be found in the EU as well.

Yet, European producers continue to adopt a complementary approach to Chinese supply, while often still relying on Chinese demand. A thorny issue for the European Commission is that, just as it is getting close to concluding its investigation on imports of Chinese battery-powered vehicles, key European carmakers are relying on Chinese production for some of their own low-end electric vehicles for sale on their domestic market (Renault via Dacia, Stellantis) or even some high-end ones (BMW), while collaboration with Chinese groups in battery production, including on European soil, is continuing. It is quite telling that the European Commission’s investigation was launched “ex officio” and not as the result of a complaint from an EU producer. A difficulty for the EC will lie in the fact that there is at least one European company among the top 15 recipients of subsidies from the Chinese government for developing battery-powered cars according to a paper by the Rhodium Group (see link here). Besides, they would have to consider the retaliation risks from China, with potentially significant adverse consequences for carmakers – in particular in Germany – for which the Chinese market remains critical.

Technically, the Commission could, depending on the results of its investigation, propose “countervailing measures” offsetting the impact of Chinese subsidies on the EU’s competitive position on this specific market. Yet before they are implemented a qualified majority of the European Council, 15 member states with at least 65% of the EU’s population could oppose them. Germany has already voiced its unease, but we note that at least one government which on other issues is aligned with populist movements (Hungary) is strongly opposed to such approach given its increasing interest in Chinese investments on its territory. The latter illustrates some of the internal contradictions of the European populists who are generally sceptical of free trade. Anyway, the European Commission is aware of these sensitivities, which would normally incentivise it to be “modest” in the quantum of measures they will put on the table.

Although it is a remarkably fine line, we could see a “cautious approach” laid out by the Europeans in their handling of China which would put them at odds with the US should Donald Trump win in November, avoiding a “blanket approach” and using the threat of targeted measures to snatch concessions from Beijing. Indeed, for China, a direct trade confrontation with the EU on top of the US could be a tall order at a time when the country is dealing with stubborn softness in domestic consumption. Finding some middle ground with Brussels could be tempting. Looking ahead, a question for Europe would lie in its capacity to maintain a cooperative line with Beijing if the US demands an alignment on an all-in “Chinese containment” strategy after November in exchange for the continuation of strategic support within NATO.

Meeting by meeting

Christine Lagarde was so focused last week on keeping a veil on the next steps that she refused to even say unequivocally that last week’s cut was the beginning of an easing process (or “restriction removal process” which better fits the ECB’s current mindset). She said this is “a strong likelihood” but gave herself some wiggle room. We suspect a key reason Christine Lagarde could not provide forward guidance on the ECB trajectory last week is a lack of consensus within the Governing Council. Before the meeting, some members explicitly laid out their stall for a gradual pace of easing - getting sometimes very specific like Klas Knot and his “one cut per quarter” proposal - while others, e.g., Villeroy de Galhau, did not want to rule out back-to-back easing. Data dependence is a nice way to “let circumstances decide”, and Christine Lagarde made it plain that on this, unanimity was reached.

Some rhetorical acrobatics were however necessary. Indeed, the ECB President highlighted the stability of the institution’s forecast for inflation by end 2025 over several batches as a key source of confidence for the ECB on the need to cut at this stage. This is very different from the very critical stance she took on the forecasts only last year. We see some very good reasons for this: models do not usually deal well with the consequences of external supply-side shocks. When they start fading and the inflationary process is driven by the endogenous, usual workings of the economy, their reliability improves.

But surely, if the level of confidence in the inflation trajectory improves, then the central bank should not have too much of an issue with disclosing a conditional trajectory for policy rates. We discussed this in detail last week and we don’t want to re-engage in this scholastic dispute now, but a fair description of the ECB’s disposition today is probably that they are quite confident they will deliver on the “end point” - inflation back to 2% at the end of the projection horizon - while removing some restriction, but they are still unsure of the right dosage. We note that their forecasts - which use market pricing as the technical assumption for interest rates - are consistent with a return to target with only limited rate cuts (3.6% for the 3-month EURIBOR in 2024 and 2.8% in end 2025, 20 bps and 40 bps higher than in March). They probably consider that their margin for error is low, especially given the higher-than-expected starting point for inflation in the projections and the upward revision in their forecasts for both headline and core CPI for 2024 and 2025.

Yet, we do not think there was a total absence of forward guidance in Christine Lagarde’s Q&A last week. Although she said the central bank did not necessarily want to see another batch of forecast before deciding another move, the insistence on the need to assess the inflation trajectory and to gather as much data as possible, plus the point on being unable to provide the market with more guidance “ until much later of the summer” in our view makes a July cut very unlikely, in line with the market’s (and our) expectation before the meeting (for the record we have been calling a June cut since September of last year).

Market pricing (see Exhibit 1) has not changed much after last week’s press conference, with still around a 2/3 probability of a total of 75 bps cuts by December 2024 - our baseline. The level of conviction remains relatively low, echoing the central bank’s own prudence. Data dependence is potentially a source of volatility since the market can over-react to every single data release between two meetings. This may be compounded by the fact that the Fed also is likely to remain in data dependent mode in the next few months.

The US dataflow remains hard to read

The market pricing for the Fed changed again drastically last week in response to a higher-than-expected print for US job creation in May according to the Establishment survey (+272K, against a consensus of +180K), accompanied by faster wage gains (+4.1% on a 3-month annualised basis, sharply up from 3.0% in April). Before the release, two rate cuts were almost fully priced by December (48 bps), with 22 bps in September already. After the release, the market was pricing only 14 bps of cuts in September and a total of 34 for December. ​ But rather than directionality, it is more confusion which in our view is the main “message” from the recent US dataflow.

Indeed, at the same time the Household survey reported a significant drop in employment by more than 400k, coinciding with a rise in the unemployment rate to 4.0%, its highest level since January 2022. We think both surveys present such shortcomings that assessing the labour market is particularly difficult at the moment. Indeed, it is very likely that the Bureau of Labor Statistics’ adjustment for birth/deaths of businesses is for now artificially boosting job creation numbers…but equally it is likely that the Household survey and its overly stable sample understates the effect of immigration on employment dynamics. It is tempting to rely on the employment component of business confidence survey to get a better sense of the “labour market pipeline” but even there, messages can be confusing. For instance, the employment component of the ISM services survey has been in contraction territory since the autumn of last year, but with a high level of volatility and in contrast with the headline reading (see Exhibit 2). Meanwhile, job openings and the “quits rate” have returned to their pre-Covid level (see Exhibit 3) but the question is of course whether such normalisation is enough to bring inflation swiftly back to target or if a phase of sub-par labour market dynamics will be needed.

In such “data fog”, the Fed’s forecast of its trajectory is likely to be the point of focus of this week’s FOMC meeting. We think the “median FOMC voter” will pencil in two cuts in 2024 – which happens to be our baseline - down from three in March. There is of course a debate around the possibility only one cut would be kept in the frame, but we think this would send too stark a message, in clear that the Fed has given on cutting before the elections (a lone cut for 2024 would be interpreted as “no easing before December”), which we think would leave the Fed with too little optionality. While time is running out, we still see the possibility that the dataflow will clarify enough in the summer for allowing the central bank to start removing some of the restriction come September.