De sprong wagen

De ECB zal deze donderdag hoogstwaarschijnlijk de rente verlagen, hoewel de dataflow de laatste tijd niet per se behulpzaam is geweest. Wij denken dat centrale banken moeten wachten tot alle indicatoren op één lijn liggen voordat ze een andere versnelling inschakelen, maar alleen in een aantal zeer specifieke gevallen: wanneer de langetermijninflatieverwachtingen zijn verankerd, wanneer de transmissie van het monetaire beleid is belemmerd, wanneer het monetaire beleid wordt tegengewerkt door het begrotingsbeleid en tot slot wanneer er sterke aanwijzingen zijn dat de neutrale rente is gestegen. Wij denken dat geen van deze voorwaarden momenteel van toepassing is op de eurozone.

Met een groot vertrouwen in het feit dat het monetaire beleid restrictief is en ook na een eerste verlaging met 25 basispunten aanzienlijk restrictief zal blijven, zou de ECB volgens ons gemakkelijk het risico kunnen nemen om haar traject voor de rest van 2024 uit te stippelen. Wij denken echter niet dat ze dat zullen doen. Wij verwachten donderdag weinig vooruitsturing, met name omdat de meningen binnen de Raad van Bestuur te ver uiteen blijven liggen om in dit stadium een consensus te bereiken. Wij verwachten dat de ECB ook in september en december zal verlagen. Een dergelijk gematigd tempo zou gerechtvaardigd zijn door enkele gevallen van weerstand tegen desinflatie in de dienstensector en het risico dat het winstmargegedrag van het bedrijfsleven de convergentie naar de doelstelling van de ECB vertraagt. Dergelijke “margeweerstand” kan echter alleen werkelijkheid worden als er sprake is van een fatsoenlijk cyclisch klimaat. De indicatoren voor het ondernemersvertrouwen zijn de laatste tijd vriendelijker voor de eurozone, maar wij denken dat de risico's nog steeds neerwaarts gericht zijn en dat de Raad van Bestuur bereid moet blijven om de restricties sneller op te heffen. De afhankelijkheid van gegevens moet beide kanten opgaan.

Omgekeerd is in de VS volgens ons voldaan aan de voorwaarden om te wachten tot de indicatoren bijna perfect op elkaar zijn afgestemd, wat de terughoudendheid van de Fed om een tijdlijn voor een verandering van houding te bespreken volkomen begrijpelijk maakt. Wij denken echter dat de dataflow precies op één lijn begint te komen, met bijvoorbeeld meer tekenen dat de consumentenbestedingen afzwakken. De arbeidsmarktcijfers van deze week zullen vanuit dit oogpunt natuurlijk cruciaal zijn, met het vacaturecijfer op dinsdag en de loonlijsten op vrijdag.

U leest hieronder de volledige analyse van Gilles Moëc.

When is the case for full data dependence strong?

The concept of “data dependence” has recently been associated with central banks adopting a very prudent approach. The opposite approach, “forward guidance”, is very rarely unconditional, of course, but it is indicative of a central bank which is quite comfortable with a certain trajectory because, based on its own experience, it can safely anticipate the effect of its past policy choices on the inflation path even if some aspects of the current dataflow are not exactly convergent. When conversely a central bank explicitly states it will make its decisions “meetings by meetings”, seemingly waiting for every single indicator to align, it implicitly conveys a sense of unease about its own policy calibration. In a nutshell, the central bank tells us that it cannot be sure its current stance is appropriate, i.e., that it will trigger its expected effect on the economy and consequently on inflation. It Is then preferable to check with as many indicators as possible that the expected effects are indeed materialising before changing course. The drawback of such “Thomist” approach is that monetary policy may not be reactive enough. Indeed, if one wants to wait until the effects of a policy fully materialise before changing it, it is highly likely that the stance will have already been inappropriate for quite some time given the transmission lags. In the face of this limitation full data dependence should be reserved for very specific situations.

A first case for us would be a situation in which inflation expectations are no longer anchored. In such case, assuming corporate and households’ price expectations are mostly adaptative – i.e., extrapolate from recent developments – an observed, lasting change in actual inflation – with even possibly an overshooting - is probably needed before the central bank can safely lower its guard. Note that this should be symmetric: it would apply both when inflation has been significantly exceeding target for a while and economic agents are positing a continuation of such inflation drift, and when the economy has been in a deflationary trap. The latter would explain, for instance, why the Bank of Japan allowed inflation to significantly exceed target again, for several quarters, before exiting its negative rate policy.

A second case would be a situation when the central bank’s efforts are counter-acted by another leg of economic policy, in clear where the fiscal stance is at odds with the monetary stance. It could make sense to wait to see “which one is winning”, i.e., what is the net effect on the real economy and inflationary pressure, before changing tack.

The presence of external shocks, or internal defects to monetary transmission, would be a third configuration in which full data dependence would be warranted. Supply-side bottlenecks can blunt the impact of the monetary stance on inflation, or an inability of the banking sector to pass the signals from monetary policy – for instance because its own balance sheet is impaired or, conversely, because ample liquidity and/or a relaxed attitude to risk is making banks ready to expand credit supply despite policy rates moving up. The central bank’s models would be far off the mark in those circumstances, justifying a sole reliance on observed data.

Finally, doubts as to where the neutral interest is, and hence where what the current level of restriction or accommodation stands. Indeed, if the central bank suspects that the neutral rate has shifted, then relying on experience to get a sense of whether “enough is enough” and it is time to think about a policy change can be misleading.

We think three of these four conditions currently apply, to varying degrees, to the United States. Monetary policy efforts have to deal with a still very accommodative fiscal policy, which under the guise of industrial policy is pushing non-residential investment up despite the rise in interest rates. True, the big supply-side shocks triggered by the pandemic and the subsequent reopening have been absorbed, as reflected in the normalisation of delivery times, but while there is no strong sign that the banking sector is failing to pass the monetary policy signals to the final borrowers, policy transmission is however hampered by the fact that, due to a past lengthening in corporate debt maturities, the bulk of the “refinancing shock” will materialise only next year. At the same time, as we have already discussed in Macrocast, residential investment has so far been protected by the reluctance of homeowners to move and lose the benefit of ultra-low mortgage rates, pushing those who had to move to new builds, blunting the impact of the rise in interest rates. Finally, it is easy to build the case for a higher neutral rate in the US, as stronger productivity gains and high immigration may have lifted potential growth.

The only condition which we think cannot be easily found in the US at the current juncture is a de-anchoring of inflation expectations. ​ True, the 5-year ahead inflation expectations reflected in the University of Michigan consumer survey still stood at 3.0% in May, one standard deviation above the 2000-2019 average, but market-based expectations derived from the pricing of inflation-indexed Treasury bonds have fully normalised: at 2.3% they are well within the Fed’s target given the usual gap between the Consumer Price Index (CPI) and the Personal Consumption Expenditure deflator (PCE) which the central bank favours (see Exhibit 1). Yet, anchored inflation expectations are a necessary condition to ease, but certainly not a sufficient one. Given the uncertainty on the other three conditions we have sketched out, that the Fed proceeds carefully and keep all options open – as reflected in the latest minutes of the FOMC – is fully understandable. Our own view is that the data flow will provide enough confidence to the Fed to cut by September (more on this in the last section of this note), but that they want to see this data flow first before sending strong directional signals makes sense.

We think that, conversely, none of these four conditions really fit the Euro area’s current configuration. Let’s start with inflation expectations. Contrary to the US, there is no time series with a long history telling us directly what is the inflation rate which is expected by private economic agents. The ECB has created such survey in 2020 only, i.e., when the pandemic was in full flow. Hawks probably regret that the median inflation rate expected three-year ahead, at 2.4%, is still above the ECB’s target, but it has been falling significantly from a peak at 3.0% in 2023 and a few more months of actual disinflation should do the trick (see Exhibit 2). Moreover, market-based inflation expectations are also consistent with the ECB’s target. There is no conflict with fiscal policy: national government have been running down the income-support schemes they had launched at the peak of the energy crisis and the overall fiscal stance is now restrictive across the core countries in the Euro area (the picture is less clear-cut in the periphery given the disbursements of the NGEU funds). Given the weak volume of credit origination in the Euro area, it can hardly be argued that the banking sector is not responding squarely to the monetary policy signals. Finally, instead of the productivity boom currently enjoyed by the US, productivity in the Euro area continues to stagnate. Potential growth, and hence the neutral rate, is unlikely to be higher today.

Accordingly, we not only think the ECB can easily “take the plunge” and cut this week despite a more complicated dataflow recently, but also that it could afford to be fairly transparent on its future trajectory beyond the rate cut for this week which they have already telegraphed, both hawks and doves. If there is enough confidence to cut in June, there should also be enough confidence in how monetary policy is operating and, of course conditional on the absence of surprises, the Governing Council could tell roughly where they would be heading to for the remainder of the year. This would have the benefit of anchoring expectations on a European market which has to deal with volatility in the US and nailing the point home about the ECB’s capacity to decouple from the Fed on a lasting basis. Yet, since our job is to predict what the central bank will do and not what we think they should do, we think they will avoid this approach and remain on Thursday very vague on the next steps after delivering their first 25 bps cut.

The ECB’s hesitations

Banque de France Governor Villeroy de Galhau highlighted an internal contradiction in the current call from the hawks to opt for a measured pace. Indeed, they are already arguing for skipping the July meeting and advocate another cut in September only. He observed that this has solidified in the market the expectation that the ECB will cut once every quarter coinciding with a new batch of forecasts. His point is that there is a contradiction between pledging “data dependence” and indicating today a certain “prudent” path. In any case, this discussion highlights the depth of the disagreements within the Governing Council, with the option of “back-to-back” cuts open to some, and already closed to others. In such configuration, providing some measure of forward guidance this Thursday will be difficult.

We do not think the slightly higher-than-expected inflation print for May which came out last week is a strong piece of evidence in this debate. Indeed, even though the year-on-year change rose from 2.4% in April to 2.6%, with the same 20 bps acceleration for core to 2.9%, driven by services prices, the volatility in the short-term momentum is strongly suggestive of “one-off effects” blurring the picture (see Exhibit 3). Just like the rebound in negotiated wages for Q1 came from an idiosyncratic effect in Germany – one-off payments to civil servants – the May CPI print for the Euro area as a whole was pushed up by a base effect in the German contribution (the disappearance from the year-on-year calculation of a drop in the price of train tickets last year). We do not have at this stage the complete dataset which would allow us to directly calculate the impact of Germany, but the data for services prices in the national measures of the consumer price index suggests that indeed, there was no sign of renewed tension across the monetary union (see Exhibit 4). Christine Lagarde in April already had made it plain that the central bank was expecting some fluctuations in inflation in the months ahead and would be ready to tolerate this. This point was made again by the ECB’s Chief Economist Philip Lane in an interview in the Financial Times before the May CPI print came out.

Yet, a key aspect of Philip Lane’s interview – which we find all the more interesting that he is not a hawk – was his insistence on monetary policy having to remain restrictive “all year long”, even though “within the zone of restrictiveness we can move down somewhat”. As such, this would not preclude the ECB from cutting rates quite quickly – e.g. at every meeting – since at 4% it is highly likely there is quite a distance from the “neutral rate” – but even if Lane did not rule out the possibility of embarking on back-to-back cuts, his insistence of the fact that inflation has not yet normalised was suggestive, in our opinion, of a measured pace of restriction removal.

We do not think the new ECB forecasts will provide a clear clue to the steps ahead. On the real economy, since March, data has broadly come out in line with the ECB’s expectations, maybe except for Q1 GDP which came at +0.3% while the ECB pencilled in +0.1%. The stronger carry over should boost euro area growth to something around 0.7% (+0.2pt) for 2024 if the March pace is maintained, but we see no reason why the ECB would change the trajectory for 2025 and 2026. On inflation, we are also expecting only small changes, tilted to the upside, to reflect the latest numbers. Core inflation was 0.1 pt higher in Q1 at 3.1% yoy and is currently running 0.2pt above in Q2. We believe the ECB should keep the same disinflation pace but that still implies an uptick of 0.1-0.2pt for 2024 core inflation (most likely at 2.7-2.8%), but we do not expect changes to the 2025 and 2026 forecasts, which would stay at 2.1% and 2% respectively.

The reason why our baseline remains that the ECB will cut only once a quarter this year (June, September, and December) is because we focus on “pockets of resistance” against the overall disinflation process in the Euro area. Even if the May print was in our view not conclusive, as we discussed earlier, we continue to see some robust price pressure in the services sector. Exhibit 5 illustrates the stark contrast in businesses’ selling price expectations across manufacturing – where they have fully returned to their long-term average – and services where, although going in the right direction, they still stood in May at two standard deviations above their long-term average (see Exhibit 5).

Price behaviour in the services sector is very sensitive to labour costs. We have expressed last week why we did not think the Q1 print for negotiated wages really puts into question the overall deceleration – and Philip Lane’s points on this issue in his FT interview suggest that this is a shared view in Frankfurt – but margin behaviour is another important term in the equation. Our colleague Hugo Le Damany produced a very thorough not on this very subject last week (link here). Margins are difficult to assess in real time, but one of his key messages is that looking at proxies – be it business surveys, or the consensus of financial market analysts for expected earnings – still point to a level of resilience in margins which could even exceed what we think was implicitly the ECB’s assumptions in their March forecasts.

Such “margin resistance” could however materialise only if decent cyclical conditions prevail. Business confidence indicators have been kinder to the Euro area recently, but we think the balance of risks still lies on the downside. We still have not really tested the Euro area for the reversal in the fiscal stance. On the external demand side, we note that the rebound in activity in China is still largely export-driven with little sign of a proper recovery in consumption, which could be major issue for some European players, while the US demand is showing signs of softening. In a similar vein, financial analysts may be too optimistic about the real economy for the remainder of 2024. In our opinion, and given your humble servant’s usual pessimism, beyond our baseline of one cut per quarter, the Governing Council should remain ready to remove restriction faster. Data dependence should go both ways: it can’t always be a reason to be prudent. It should at times be a signal to be bold.

Meanwhile, in the US….

We expect the US commentariat on Thursday to focus on the ECB’s “decoupling from the Fed” but in our baseline it’s more a lag than a proper decoupling. There is still no smoking gun of course, but we continue to see more reasons to hold to our baseline that the Fed will be able to cut in September. The April reading for the core Personal Consumption Expenditure deflator (PCE) confirmed the improvement reflected in the CPI version of inflation a few weeks ago. Core inflation stabilised at 2.8% in year-on-year terms (it was 2.75% to be precise) but the recent momentum softened noticeably after 3 months in a row of re-acceleration (see Exhibit 6). Inflation in core goods continues to edge higher – a normalisation after the collapse in demand for goods when the economy reopened from Covid, combined with the exhaustion of the impact of the dollar appreciation in imported goods – but crucially, the deceleration in services is more than offsetting this (see Exhibit 7). 

This piece of news came together a lower-than-expected print for private consumption in April. The monthly decline was slight (-0.05%), but together with the small downward revision to the March reading (from +0.5% mom to 0.4%), and a somewhat larger revision for the whole Q1 (from 2.2% qoq annualised to 2.0%), this adds to the sense that this essential engine of US growth is starting to sputter.

This week’s dataflow will provide some more interesting readings into the current health of the US economy. We will take a hard look at job openings, on Tuesday, for confirmation that the US labour market is cooling. Interpreting the JOLTS numbers is not always easy though. While openings have been falling quite regularly since a peak in March 2022, they remain higher than before Covid. The quits rate is probably more indicative. It has fallen below the late 2019 level since November of last year. A further decline would be good news for softer wage dynamics. Yet, the market will of course likely focus on the payroll for May, released on Friday. ​ The Bloomberg consensus – as we write these lines – is on a stabilisation at c.170K. As we argued last week, the NFIB survey points to some weakness ahead.

Even if it materialises in this week’s print, we would not expect a quick change in the Fed’s rhetoric on this side of the summer, but we continue to hold the view that “it would not take much” to tilt the scales towards a macro configuration consistent with a swifter convergence towards the central bank’s inflation target, opening the door to a rate cut in September.