De landing zacht houden

De keuze van de Fed om haar versoepelingsproces te beginnen met een verlaging van 50 basispunten is gedurfd, maar de nieuwe prognoses die bij het beleidsbesluit horen, maken duidelijk dat er volgens het FOMC in totaal 200 basispunten aan verlagingen nodig kunnen zijn om “de zachte landing zacht te houden”. ​ Terzijde: het FOMC is van mening dat het, waarschijnlijk deels omdat het ervoor heeft gekozen om “met een knal te beginnen”, de beleidsrente in deze cyclus niet naar een echt accommoderend gebied hoeft te brengen. We denken inderdaad dat het geen toeval is dat het niveau waarop de Fed Funds-rente eind 2026 belandt, samenvalt met hun nieuwe schatting van het “langetermijnniveau”. Wij denken dat het een belangrijke aanwijzing is die de obligatiemarkt niet mag missen.

De prognoses van een centrale bank moeten meer worden gezien als een “intentieverklaring” dan als een echt “actieplan”. Achteraf gezien was de dot plot van juni te optimistisch en reageerde deze te sterk op de opleving van de diensteninflatie begin 2024. Omgekeerd zou de plot van september wel eens overdreven kunnen reageren op de teleurstellende loonstrookjes van de zomer. Om de kans te meten dat dit laatste gebeurt en om het contrast met de ECB te karakteriseren, kijken we naar enkele van de ontbrekende gegevens in de beslissingsfunctie van de Fed. Veel zal afhangen van de uitslag van de verkiezingen in november, en wij denken dat de Fed zich misschien moet terughoudend opstellen wat betreft de hoeveelheid verlagingen die ze nog moet doorvoeren als Donald Trump wordt verkozen. Omgekeerd denken wij dat de vooruitzichten voor de eurozone duidelijker zijn. De enige factor die het optimisme zou kunnen ondersteunen, zijn de aanhoudende ongunstige ontwikkelingen op het gebied van het arbeidsaanbod, die tot uiting komen in de nog steeds hoge aanwervingsmoeilijkheden in de bedrijfsenquêtes. We kwantificeren de impact hiervan op de loonontwikkelingen en komen tot de conclusie dat deze een zichtbare rol, maar geen cruciale rol, hebben gespeeld in de loonontwikkelingen van de afgelopen twee jaar, waardoor de deur wordt opengezet voor een voortzetting van de loondaling ondanks een nog steeds beperkt arbeidsaanbod. Gezien de externe omstandigheden en het vooruitzicht van bezuinigingen op de begroting volgend jaar - die qua richting zekerder zijn dan in de VS - zou de ECB wel eens gedwongen kunnen worden om haar versoepelingsinspanningen te versnellen. We zouden dan weer terug zijn bij het “spiegelrisico” dat we begin dit jaar benadrukten: de mogelijkheid dat de Fed uiteindelijk “te veel” zou doen en de ECB “te weinig”, althans in eerste instantie.

Fed shifting focus

The Federal Reserve (Fed) choosing to "rip the band aid" and cut by 50bps sends a strong signal: this is the first time in the US modern monetary policy history that it resorted to a “jumbo move” without an obvious recession and/or financial crisis brewing (the last two instances were 2001 and 2007). ​ The effect is in our view even stronger because the decision was nearly unanimous. True, it was not completely a walk in the park for Powell, as Michelle Bowman’s dissent in favour of a 25-bps increment was the first one of a Fed governor since 2005, but this hardly constituted a mass rebellion, suggesting the new stance reflects solid determination across the Federal Open Market Committee (FOMC).

The rationale was straightforward: Powell stated at the very beginning of the press conference that the Fed is equally focused on its two objectives: price stability and full employment. As the FOMC now clearly believes that the battle to bring inflation back to target is about to be won, they can shift their focus on their second objective. ​ Their sense of concern on the latter was plainly illustrated by the new forecasts: while the projection for GDP growth has not changed, staying at a slightly above potential 2% in 2025 and 2026, the associated unemployment rate was taken up by 0.2 pp in 2025 (4.4%) and 2026 (4.3%) relative to the June batch. That would still qualify as a very soft labour market landing by historical standards, but it also means that (i) they take the recent spike seriously, and (ii) that they do not think there is a need for a massive additional deterioration in unemployment to bring inflation back to 2%. There was also a small downward revision of core inflation in 2025 to 2.2% from 2.3%. We can reverse the sense of the equation: it is precisely because the Fed is now expecting a substantial monetary easing that they can maintain a decent GDP trajectory ahead. As Powell put it, they "did not want the Fed to fall behind" what's needed to deliver on the full employment objective.

Indeed, while in the Q&A Powell chose to keep it sober on the next few moves, mentioning that the Fed "is not in a rush" and will make its decision "meeting by meeting”, the new dot plot is consistent with 50bps worth of additional cuts in 2024, and a 100 more in 2025 to 3.4%. This is still higher than what the market is pricing now (2.85%) but only one “normal” rate cut away from what the market was still pricing at the beginning of September (3.15%). So, we have a very pre-emptive and potentially bold Fed, which is not trying to stand in the market’s way. We are not sure such boldness was warranted by the data, but again this a strong message. 

Governor Waller in a candid interview after the FOMC decision also sent clear signals: he would support 25 bps cuts “if the economy continues to look fine”, with the possibility to resort to more aggressive moves “if the data is soft and continues to be soft”. While, to be fair, he also mentioned the possibility to pause if need be, it is quite telling that even in a benign environment, regular easing would be needed. ​

Now, as usual the actual calibration of the trajectory may end up quite differently from the dot plot, which has never been a very good predictive tool. The recent gyrations in the “dot plot” have been significant and they were “sequentially discontinuous”, in the sense that while the overall direction of travel – easing ahead – was clear, there was no regular shift towards a more hawkish or more dovish trajectory. As exhibit 1 illustrates, June marked a “hawkish recalibration” relative to March, with September turning hard to the dovish side. Arguably, both moves could be seen as a “knee-jerk” reaction to the recent data flow. June was the height of the inflation rebound theme, with services prices re-accelerating, amid signs of lingering pressure on the labour market. Symmetrically, the September dot plot may result from an over-reaction to the soft payroll readings over the last two months.

Another key point however for us is that the FOMC implicitly believes the policy rate won’t have to go into properly accommodative territory to deliver a soft landing to the US economy. Indeed, the long-term level of the Fed Funds – which is routinely taken as a proxy of the FOMC’s estimate of the “neutral rate” – now stands at 2.9%, up 0.1pp from June (and now visibly above the 2.5% level where it had been standing for a long time before the Covid crisis). In 2026, the “median FOMC member” sees the Fed Funds rate landing at this very level. The underlying message is that merely removing restriction will suffice to keep the economy in line with potential and bring inflation back to target at the same time, possibly because such removal is starting early, and is proceeding, at least at the beginning, at a fast clip. This is something which bond market participants may want to “meditate”. If the dot plot materialises, then, unless one wants to make extreme gambles on an eternal compression of the term premium, there is little reason why 10-year yields should fall much further, while a still large supply of treasuries could push long-term yields back up in the years ahead.

Mirror hockey sticks across the Atlantic?

Of course, a “dot plot” is statement of intent rather than a proper plan. Yet, such statement contrasts quite starkly with the ECB’s extreme caution. To explore this, we want to return to the views we expressed at the beginning of 2024. At the time, we warned against the existence of opposite risks across the Atlantic. In the US, evidence that the monetary policy stance was cooling down the economy fast enough to bring inflation back to control was missing then, which made us question the market’s expectations of aggressive easing starting in the first half of the year, while in the Euro area our concern was that the European Central Bank (ECB) might have gone a bit too far on its restrictive stance ​ – we diplomatically wrote that the September 2023 hike was “not absolutely necessary” – and it was at risk of ignoring for too long bad signals from the real economy. Nine months later, and now that the two central banks have started to ease their stance, we want to take stock again of the balance of risk across the Atlantic.

On the US side, we were right not to expect early easing by the Fed – it even came later than we thought – but we were wrong on the kind of macro configuration it would take to really get inflation back under control. Indeed, the good news is that the “hump” in consumer prices seen in the winter of 2023-2024 – amid still strong cyclical conditions – dissipated by late spring-early summer without a massive GDP slowdown and painful labour market correction. Soft landings are to some extent “miraculous” – it takes a very delicate set of ingredients to get there – but it seems that it is what we have now. The labour market is normalising while still allowing decent job creation (around 1% on an annualised basis) and the Atlanta Fed’s nowcast, after last week’s confirmation in the retail sales for August that consumer spending is holding up, is back to a more than decent 3%. Judging by its effects on the economy rather than by reference to an always elusive “neutral rate”, it is not even obvious that the current Fed stance is that restrictive: credit production did not collapse, and the housing market correction has been very soft.

In such circumstances, the choice for the Fed is not as straightforward as what the nearly unanimous decision last week suggests. True, if a mere soft landing has been able to trigger significant disinflation, there is a danger in maintaining a too restrictive stance for too long: the soft landing could easily turn into something harder, closer to “proper recession”, sending inflation below the Fed’s target again. Yet, symmetrically, the decent state of the economy offers space to the Fed to continue easing at a prudent pace for now, carefully calibrating each step depending on the dataflow, and declaring victory on inflation too soon remains a distinct risk.

To make a choice between these two avenues, there is a key piece of information missing for the Fed: the contribution from fiscal policy into 2025 and 2026, which itself depends on the outcome of the elections in November. Kamala Harris has not been very precise in laying out her fiscal plans. Our understanding is that she would allow only a fraction of the 2017 tax cuts to expire, preserving most of the US middle class, the tax hikes thus affecting only those with the lowest propensity to spend. On the spending side, the stimulus effect of the Biden era’s IRA and Chips Act would remain largely untouched. This could result in an overall neutral to slightly expansionary fiscal stance. There is also some vagueness on Trump’s side, but significant tax cuts designed to offset the adverse effect of tariff hikes on purchasing power would be very likely. Fiscal policy would then become significantly expansionary, at a time when mechanically the trade war would lift US inflation.

Given this crucial uncertainty, it would make sense for the Fed to suspend judgement on the quantum of monetary accommodation which will be needed next year until the dust settles on the political landscape. ​ To complicate matters further, this goes beyond determining the victor of the presidential race. The balance of power in Congress will be a crucial input. This is why we have a “problem” with the current aggressive market pricing for the Fed’s trajectory; investors are – or should be – in the same situation as the FOMC, irrespective of the current message from the “dot plot”. As long as political uncertainty persists, we don’t think it is wise for market participants to take such a binary approach.

We think the level of uncertainty is lower in Europe. First of all, it is difficult to argue that inflation cooled down in the Euro area within a “soft landing” configuration. It rather seems that Europe is stuck in a sort of never-ending incapacity to take off. The Euro area has been repeatedly brushing with recession since it exited from the pandemic. Even when taking the gap in potential growth into account, Europe is much closer to a nasty downturn than the US. We noticed last week that in its last forecasts the ECB shaved its GDP projections by only 0.1% and kept the list and characterisation of risks unchanged relative to June, which we found surprising given the recent dataflow. There is also much more evidence in the Euro area than in the US that the monetary stance is indeed restrictive: credit origination has taken a hit, and business bankruptcies are now rising fast, more than offsetting the Covid lull.

True, an ingredient from past European downturns is missing: the labour market has been resilient so far, with the unemployment remaining at historical lows with none of the uptick seen in the US over the last few months. Supply-side conditions differ. While strong immigration is lifting the population ready and willing to work in the US, the Euro area is dealing with a dearth of immediately available workers. This could trigger lingering pressure on wages and hence impair the continuation of the disinflation process. ​ This is however contradicted by the recent hard data – pay per head decelerated in Q2 – and, looking forward, by the available surveys.

We think it was an important signal that hiring difficulties, among the list of factors mentioned by businesses to explain their inability to raise production, has been taken over by weak demand for 2 quarters now (see Exhibit 2). A “hawkish observer” could however argue that the two factors could combine to trigger a “stagflation configuration”, with mediocre – even negative – economic growth co-existing with still strong inflation. Indeed, when correcting the balance of opinion on the various “production bottlenecks” factors for their long-term average and historical volatility, demand constraints have merely normalised, with a z-score very close to zero, while the labour constraint is still at two standard deviations above its long-term average (see Exhibit 3).

We want to take a more quantified approach to this issue by injecting in a canonical econometric model of wages per head – with lagged inflation and productivity as explanatory variables – the balance of opinion on hiring difficulties. Because of the gyrations at the time of the pandemic we also introduced two dummies (one for 2020Q2, the other for 2021Q2). ​ As we expected, the “hiring difficulties” variable comes out as a statistically significant, suggesting that, indeed, the dearth of manpower magnified the impact of the initially supply-driven inflation shock to propel wages in Europe over the last two years. We can use our model to calculate the contribution of hiring difficulties to the yoy change in wages (see Exhibit 4): it was visible, but not crucial. The difference in contribution relative to the 2018-2019 period when inflation was still tame but hiring difficulties were already a prominent topic in the European debate, is relatively small – about 0.6 percentage point. The main contributor to the massive acceleration of wages in Europe was a catch-up process with an inflation shock which was essentially externally driven. Once the external shock subsides, wages would maintain for a while a robust growth rate, given Europe’s specific labour market institutions (dominance of collective bargaining, automatic indexation of some national minimum wage schemes) but ultimately, we do not think the supply side difficulties on the European labour market could, on their own, impair much the ongoing landing of wages. ​

All in all, and also taking on board the restrictive tilt of fiscal policy in Europe next year – the quantum can be discussed, but not the direction of travel, a key difference with the US as long as the election is not settled – we think the risk of inflation ultimately undershooting the ECB target and falling back under 2% remains significant and should be met with a clearer commitment to an easing trajectory by the Governing Council. On the margin, the Fed’s decision last week should make such “dovish conversion” by the ECB easier, since it reduces the risk the euro would depreciate significantly if the pace of rate cuts accelerated in in the Euro area. ​ Yet, judging by the state of the debate at the Governing Council. It may take time for such conversion to materialise. In the end, we could see the monetary policy trajectory of the Fed and the ECB appear as two “mirror hockey sticks”: the ​ ECB would initially refuse to cut quickly, before being ultimately forced into an acceleration by deteriorating macro conditions and inflation falling faster than in their central scenario, while the Fed would follow the symmetric “hockey stick “shape, cutting initially quickly before being stopped out by a rebound in inflation triggered by expansionary fiscal policy and trade tariffs. ​ ​