CITI’S TAKE (10 Jun 2026 03:14:37 ET)
The Governing Council of the ECB is fully expected to hike rates by 25 bps this week. The real focus of this week’s meeting is on what communication may convey about future policy moves. We have not altered our baseline expectations of a second, back-to-back hike in July, which is a minority view. We explain here why we have maintained this modal expectation: essentially, we cannot rationally justify a pause when a Governing Council driven by an insurance motive now faces a substantially worse inflation outlook than in March.
Worse-than-adverse projections — Our “projections of the ECB staff projections” lead us to expect that the Governing Council will be presented a new baseline scenario that implies a growth outlook as bad as the “adverse” scenario put forward by the staff in March, but also an inflation outlook that is substantially worse than in that adverse scenario. At the relevant policy horizon (approximately two years) headline and core inflation projections are likely to be above target. This is despite an implicit assumption of at least two and possibly three rate hikes.
The first hike evidences an “insurance” motive — Faced with lower growth and higher near-term inflation, a central banker’s reaction function depends on whether they place more emphasis on demand management or expectations management. The fact that the ECB decides to hike rates now evidences a bias towards the latter. That is consistent with an “insurance” motive, as suggested by several Governors. An insurance-driven policy assumes however as much tightening as is required to all but eliminate the risk of persistent inflation overshooting. Given the upward revision to market breakevens, consumer inflation expectations and ECB projections at a policy-relevant horizon, we do not see how this justifies a pause in the hiking process.
Is a pause really a stop? — The question that the Governing Council may face in practice is not whether to pause, but when to stop hiking. We see a strong probability that lower aggregate demand, weaker employment dynamics, a lesser ability of firms to pass on cost increases to consumers and of course – more hypothetically – a reopening of the Strait of Hormuz convince the Governing Council that it no longer needs to guard against a diminishing risk, and that a pause in policy tightening – whenever that takes place – becomes the end of it.
ECB: once hikes are underway, what is the point of a pause?
As has often been the case over the past few years, the focal point of attention around this week’s meeting of the Governing Council of the ECB is not the decision on the day. It is what communication in the aftermath of the decision may convey about the subsequent policy moves. We have not changed our baseline expectation that if the Council is to continue hiking policy rates it is more logical to expect back-to-back hikes (and so a 25 bp hike in July) than a more staggered tightening (and so a pause in July, followed by a hike in September). Majority of investors we have talked with seem to disagree. So, in this note, we explain our reasoning. We do not pretend to have a very high conviction as to the policy path that will be followed, if only because it so heavily depends on unpredictable military and diplomatic development in the Middle-East. Our argument is simply that we have to retain as a baseline expectation a policy path that has a genuine internal logic.
Projections of the projections: lower growth again, inflation higher for longer
As a starting point, we reiterate very briefly that we still see the current shock caused by the conflict in the Middle-East and the closure of the Strait of Hormuz as a textbook negative supply shock. Its effects are therefore stagflationary, in the sense that they raise prices and lower growth. Our expectation at the start of the conflict was that higher energy prices would lower demand for energy itself, but also cause an eviction effect, reducing the ability or willingness of consumers to spend money elsewhere. As services is the sector where most discretionary consumption takes place, the inference was that it is in that sector that weakness of demand would materialise more meaningfully. And as services are relatively labour-intensive, the risk was a softening of employment dynamics, which in turn may weigh more sustainably on sentiment and consumption. So far, economic developments seem in keeping with this broad-brush description.
What matters of course for policy (in the near term at least) is not what we think but what the Governing Council thinks, and so the dynamics of demand and prices reflected in the staff projections. On the whole, we do expect the projections to be unveiled this week to convey a similar narrative to that we just summarised.
The cut-off date for the input in the projections remains somewhat uncertain to us as the ECB may again decide to delay it closer to the meeting date, like they did in March. We assume however that the old rule of freezing the technical assumptions around three weeks before will apply this time.
Since the March cut-off date, the main change to the technical assumptions is a higher short and medium-term oil price, by some 15% (see Fig 1). Gas prices are roughly unchanged (lower short-term TTF contracts, slightly higher longer-term ones), the euro is unchanged, and the market pricing for short-term interest rates is marginally higher (by around 12bp), alongside higher 10-year yields.
On balance, the technical assumptions imply higher headline HICP in 2026 and marginally higher in 2027 (a higher oil price profile is partly offset by more negative base effects).
On GDP, we assume the latest down-revision to 1Q GDP (from 0.1% to -0.2%QQ, driven solely by Irish volatility) will not be included as it became available after the cut-off date. If so, 2026 growth may be revised only slightly lower due to softer flash 1Q GDP and weakening surveys in 2Q. We think the ECB will still treat this near-term weakness as temporary and will forecast growth to return to around potential from later on this year.
The new projections will likely embed higher forecasts for both headline and core inflation. Higher core HICP will likely mirror a higher starting point in 2Q (by 15-20bp), plus somewhat bolder assumptions on second-round effects than assumed in March (based on the recent increases in firms’ selling price intentions) and higher unit labour cost growth in 1Q26. These factors will be only partly offset by lower GDP growth. Overall, we see HICP revised up by 0.4pp and 0.5pp in 2026 and 2027, to 3.0% and 2.5%, respectively, and core inflation revised up by 0.2pp and 0.3pp in 2026 and 2027 respectively, and possibly also by 0.1pp in 2028. We see 2028 HICP still seen converging close to 2% (2.1% in March).
The new baseline is worse than the old adverse scenario, which already included at least two hikes
Neither the revisions to growth or inflation projections are innocuous. On the real side, they imply that yet again the euro area fails to grow at potential and that (unless one assumes a permanent erosion of potential), the output gap is reopening. On the nominal side, we have a yet another upward deviation of inflation from target, both higher and – crucially – twice as long as was assumed in March. One should remember indeed that in the ECB staff’s scenarios published then, headline inflation was back to 2% or thereabout in 2027 under both the baseline and adverse scenarios. Only under the so-called severe scenario was inflation projected to remain much higher. Further, under the adverse scenario, inflation was projected to fall as low as 1.6% in 2028. If our estimate is correct, the ECB would therefore now face a new baseline scenario that is effectively as bad as the old adverse scenario for real developments and worse for inflation through the policy-relevant horizon.
A hike now evidences an insurance-driven policy mindset
This begs the usual question: in the face of a stagflationary shock, should the central bank put more weight on the weak outlook for demand or on the strong outlook for prices? Bypassing entirely the usual explanation about the magnitude and duration of the shock, that we sometimes think is more used as an ex-post justification of decisions than as an ex ante framework to guide those decisions, we underline again the importance of what central bankers believe produces inflation. We routinely divide central bankers between two intellectual schools of thoughts: those who fundamentally believe inflation to be a matter of balance between supply and demand; and those who believe that inflation expectations play a genuine role in the formation of prices.
Members of the former school are much more likely to see in the weakening of demand the promise of downward pressure on inflation later on; they are much more likely to show patience with respect to pace of return of inflation to target; and they are much more likely to be adverse to contributing to a more permanent erosion of economic tightening.
Members of the second school are more likely to exhibit impatience in the face of a price shock, because they will fear that higher prices raise inflation expectations, which by itself raises inflation dynamics. We have argued in the past that this also makes them more asymmetrically adverse to risks around the target. The target itself may be symmetric, but aversion for deviation from it is not. The reason is that downward nominal rigidity arguably makes a de-anchoring of inflation expectations less likely on the downside than the upside.
We do not discuss further this distinction between intellectual approaches to the formation of prices further – we did so in the past – but remain at the disposal of readers who which to discuss it further. We note however that while we have generally been skeptical of the relevance of inflation expectations for price dynamics, current developments in Japan – where one large exogenous price shock seems to have meaningfully contributed to an inflation regime shift – give us reason to question our position.
What matters for the purpose of this current discussion is more pragmatically that communication and actions by the Governing Council abundantly evidence that the majority of Governors are much closer to the second than the first school of thoughts. Indeed, the primary reason why the Governing Council seems about to unanimously decide to hike interest rates appears to be an insurance motive, i.e. a perceived urgency to guard against an upward drift of inflation expectations. If the Governing Council belonged in majority to the first school of thoughts described above, it is likely that they would see in downward growth revision the promise of decelerating inflation (without policy action) and would delay a hypothetical tightening of policy until evidence of second-round effects. It is noteworthy in this context that in the press conference of April 2026, the President of the ECB unequivocally indicated that the Council did not see any sign of second round effects, but at the same time that “directionally”, she knew where the Council was going (i.e. hike in June). That conveyed beyond doubt that the burden of proof was not on evidencing the presence of second round effects, but on evidencing their absence.
If insurance is the motive, how could just one hike be enough?
In this context, we struggle to find an internally consistent justification for a pause in July to follow the foretold hike in June. We provide below our arguments, for the purpose of debate: First, we think important that the Governing Council assumed in March (and by extension in April) at least two rate hikes, as this was the market-based assumptions underpinning the staff forecasts and scenarios then. If they assumed that two hikes were necessary to bring inflation back to target within one year, and now they face projections were – with at least two hikes assumed – inflation still does not come back to target within two years, why wouldn’t they intend to deliver those hikes as a default course? And if they intend to deliver those hikes, why would they need to pause on the way?
Further, it is legitimate to ask whether the ECB intends to set policy to bring inflation asymptotically back to target, or to go further than that. The reason we say so is that Governors have suggested that hiking policy was a form of insurance. So what does “insurance” mean in a monetary policy setting? The only consistent interpretation is that policy is calibrated to minimise the probability of the outcome that the central bank aims to “insure” against. In the current situation, that outcome is a persistence of above-target inflation that would lead to a de-anchoring of expectations. Given that the ECB cannot realistically control the variance of the distribution of plausible inflation at a two-year horizon, insurance must mean pushing down the distribution enough that the tail that remains above target is very small. By construction that means pushing the mode, the mean, and the bulk of the distribution below target.
So if the first hike responds to an insurance motive, surely it must be followed by others in short order, because again, ECB staff projections clearly suggest that one hike will not come anywhere close to delivering the sought level of protection.
That argument would collapse upon itself if a single hike were enough to cause a psychological effect, convincing consumers, firms and investors of the commitment of the ECB to such an extent that inflation expectations fall as a consequence. We are rather skeptical, because if that were the case, the well-anticipated rate hike of this week would already have had that effect.
A counter argument we are often given is that the Governing Council is reluctant to tighten policy in view of the ongoing erosion of demand by fear of having to reverse course and subsequently cut rates, as happened in 2011. That would justify a more staggered approach, with small steps (25 bp hikes on a quarterly basis) followed by pauses to take the time to assess incoming data until new projections are published.
This may turn out to be a correct view, but we have too much faith in the analytical depth of the Governing Council to make that a baseline scenario:
For one thing, if the Governing Council really adopts an “insurance” approach, reversing course would not be a failure but arguably success. As explained above, an insurance-driven policy implies that the central bank intentionally pushes the distribution of inflation too much downwards to guard asymmetrically against the outcome deemed more detrimental in the medium-term. Once this outcome eliminated, it would be entirely normal for monetary policy to reverse course. Further, we do not believe that the Governing Council needs to wait three months to have updated projections. It evidently adjusts its assessment of the economy on an ongoing basis.
The risks of a reversal of policy course strikes us also as very different now than in 2011, an episode that very few governors (other than Fabio Panetta) experienced “from within”. Back then, the ECB hiked policy rates in a context where (1) the euro area was entering a foreseeable self-imposed recession following policy decisions that contributed to curtailing access by banks to unsecured funding and triggered a credit crunch and (2) euro area governments were predictably adopting austerity policies, which aggravated the shortage of domestic demand relative to supply, and of domestic investment relative to savings, which ended up pushing inflation and real rates downwards. Now, as we have highlighted for some time, the trend seems to us to go the other way (structurally higher and more volatile inflation, trend rise in the equilibrium real rate, as also suggested by Isabel Schnabel). If we are correct, and if the Governing Council shares that view, then they should not fear too much a repeat of the policy mistake of 2011.
This is not to say that we believe that hiking rates now is the desirable policy. We remain of the view that the erosion of domestic demand means that the ability of firms to pass higher input costs to their clients is relatively limited, and that until this view is proven wrong, the case for hikes in the first place is unconvincing. But that reflects the fact that we place a lot less emphasis on inflation expectations than Governors. If we take as a working assumption that inflation expectations matter, then we do not see how a single 25 bp hike can be considered as sufficient ex ante by the Governing Council, given again the slippage in inflation projections at the relevant policy horizon. For that reason, we have chosen to maintain as a baseline expectation a second rate hike in July.
Whenever there is a case to pause, there likely is a case to stop hiking altogether
We fully recognize that this second hike may not happen. It may not happen for bad or good reasons. The bas reasons would be that the Governing Council is not really committed to an insurance approach. But then, why hike now in the first place? The good reason would be that by July, there has taken place enough demand destruction to bring support to our views that the pricing power of firms is considerably lower than in 2022, and that the risk of the current exogenous price shock morphing into endogenous inflation dynamics is remote. Another good reason – albeit a more tentative one – would be that by the summer, the military and diplomatic situation in the Persian Gulf has eased enough that the negative supply shock is on its way to being reversed. In both those cases, the Governing Council would have no reason indeed to hike rates in July, but then it would have no reason either to hike rates in September, or later this year. That is why our general expectation remains that when the Governing Council puts on pause the tightening process it is about to initiate, it really puts an end to it. Whether this happens in July, September or beyond depends to a large extent on entirely exogenous events (geopolitics), over which we have no real visibility. It also hinges on more endogenous factors, primarily the elasticity of consumer demand to prices. As we think the latter is high, we lean towards a relatively early end of tightening (after 50 bps). But in any event, we think more rational to project back-to-back hikes followed by a long plateau then to assume a succession of small hikes and pauses.