Skip to content
Back to Dashboard
Macro Outlook/European Central Bank

European Central Bank Policy Analysis

The European Central Bank is hawkish at its September 2026 meeting — it hiked 25bp to 2.50%. A second unanimous 25bp hike, to 2.50% effective 16 September, delivered exactly as the 27 August account foreshadowed: "another rate hike would likely be necessary unless the inflation outlook imp...

8 meetings available

Analysis: European Central Bank - 10 September 2026

Decision: hiked 25bp at 2.5%

Stance: hawkish (Confidence: high)

A second unanimous 25bp hike, to 2.50% effective 16 September, delivered exactly as the 27 August account foreshadowed: "another rate hike would likely be necessary unless the inflation outlook improved significantly". It did not improve - August HICP rose to 3.3% from 2.9% on a 14.3% energy surge - so the Council executed. Lagarde framed the decision bluntly: "The conflict in the Middle East continues to generate inflation pressures, and inflation is set to remain well above target for an extended period." The hawkish read is reinforced by the projection round rather than the spot data: 2027 and 2028 core were revised up to 2.6% and 2.3%, leaving underlying inflation above target at the end of the ECB's own medium-term horizon, while growth upgrades for 2026 and 2027 removed the downside-risk argument that justified July's hold. What keeps this short of maximum conviction is the composition: core eased to 2.4% and services to 3.0% in the same print, so the tightening rests on preventing second-round effects from an imported shock rather than on any domestic inflation impulse.

Direction: hiking (data-dependent)

Key Takeaway:

The September hike is insurance against an imported energy shock, not a response to domestic overheating, and the data makes that unusually stark: not one of the six standard hike conditions is fully met, while headline HICP at 3.3% sits alongside core at 2.4%, services at 3.0%, unit labour costs down to 2.6% and capacity utilisation at 78.5%, well under its long-run average. The ECB is tightening into a domestically disinflating economy because it judges the cost of second-round effects asymmetric - a defensible reaction function, but one that inverts if energy mean-reverts. The counterweight to a dovish read is the ECB's own projection of core at 2.3% in 2028, which on a medium-term-oriented framework implies 2.50% may still not be terminal. Divergence rises from low to medium and the pivot risk flips from hawkish to dovish, though the timing is a 2027 question rather than an October one.

Next Action Probabilities

Hike
Probability:medium
Hold
Probability:high
Cut
Probability:none

Inflation Assessment

Trend
ValueTarget
rising-
Concern: high

What Changed (JulSep)

Stance
Jul:Hawkish
Sep:Hawkish
Change:→ Same, but now acted on rather than signalled
Inflation view
Jul:"the full effects of the energy shock have yet to play out"; inflation "well above target into the first half of 2027"
Sep:"inflation is set to remain well above target for an extended period"
Change:↑ Hardened - horizon extended, conditionality dropped
HICP headline
Jul:3.2% → 2.8% (June)
Sep:2.8% → 2.9% → 3.3% (August)
Change:↑↑ Accelerating - the improvement July was conditioned on never arrived
HICP core
Jul:2.6% → 2.4% (June)
Sep:2.4% → 2.5% → 2.4% (August)
Change:↑↓ Flat and easing - the hike was not driven by underlying inflation
Energy inflation
Jul:8.5% (June, down from 10.8%)
Sep:14.3% (August, up from 10.3%)
Change:↑↑ The entire source of the headline re-acceleration
Growth view
Jul:"economic growth will remain modest in the near term"; Q2 only modestly improved
Sep:"greater than expected resilience of the euro area economy"; 2026 growth revised up to 0.9%, 2027 to 1.4%
Change:↓ Less concerned - removes July's reason to wait
Staff projections
Jul:June round: headline above target into H1 2027
Sep:Headline 3.0% / 2.5% / 2.1%; core 2.5% / 2.6% / 2.3% (2026/27/28) - 2027-28 revised up
Change:↑ Core still above 2% at the end of the horizon
Wage pressures
Jul:Negotiated wages 2.44% (Q2); Lagarde on second-round effects: "We are not seeing it."
Sep:Compensation per employee 3.3% (from 3.5%); unit labour costs 2.6% (from 3.5%)
Change:↓ Still cooling - the one hike condition the hawks cannot claim
Forward guidance
Jul:"The burden of proof is on data"; no September decision predetermined
Sep:"We are not pre-committing to a particular rate path"; "We are not taking a view as to which direction we go at our next meeting"
Change:→ Same data-dependence, now from a higher policy rate

Key Language Shifts:

  • -Confirmed: the 27 August account's conditional near-commitment - "another rate hike would likely be necessary unless the inflation outlook improved significantly" - was executed on schedule. August HICP rose to 3.3%, so on the Council's own test the outlook did not improve and the hike followed automatically.
  • -Modified: "the full effects of the energy shock have yet to play out" → "inflation is set to remain well above target for an extended period" → the ECB has stopped describing the shock as prospective and now treats above-target inflation as the base case it must lean against.
  • -Added: "greater than expected resilience of the euro area economy", with 2026 and 2027 growth revised up → the downside-to-growth argument that justified July's hold has been withdrawn, leaving nothing to offset the inflation case.

Key Quotes

The conflict in the Middle East continues to generate inflation pressures, and inflation is set to remain well above target for an extended period.

The rationale for the hike in one sentence, and a notable hardening from July's "the full effects of the energy shock have yet to play out". The ECB has stopped treating above-target inflation as a prospective risk and now treats it as the base case it must lean against - which is what converts a pause into a resumed tightening cycle.

We are not taking a view as to which direction we go at our next meeting.

The most important line for positioning. After two hikes in three meetings, Lagarde explicitly refused to characterise 29 October in either direction - the first genuinely two-sided guidance of this cycle. Contrast the July account, which pre-committed conditionally to a hike; that conditionality is now gone.

We don't focus on the meeting and on the data to annoy... observers. We do it because we are in this uncertainty that can change things almost overnight.

Lagarde defending meeting-by-meeting data dependence as a response to genuine shock uncertainty rather than a communication device. It signals that the ECB itself does not know whether 2.50% is terminal, and that a single energy or wage print can move the October decision either way.

We are not pre-committing to a particular rate path.

Retained verbatim from July despite the hike. The Council is deliberately refusing to convert two hikes into a guided sequence, which caps how much further tightening the curve can price without new data.

Staff projections: core inflation 2.5% (2026), 2.6% (2027), 2.3% (2028).

The quietly hawkish element of the round. Core does not return to 2% anywhere in the projection horizon and 2027-28 were revised up. On a medium-term-oriented reaction function that is an implicit admission that 2.50% may not be the terminal rate, even as the spot core print eases.

Economic Data vs CB Rhetoric

Why this matters: Central banks may downplay inflation concerns in their official statements, but economic data tells the real story. If inflation consistently rises beyond the target band, policymakers will eventually be forced to act — regardless of their rhetoric. Comparing what they say versus what the data shows helps anticipate policy pivots before they happen.

HICP HeadlineMonthly
OlderPrevLatestTrendTarget
2.8%2.9%3.3%2%
Well above target and accelerating - energy at 14.3% accounts for essentially all of the August jump
HICP Core (ex food/energy)Monthly
OlderPrevLatestTrendTarget
2.4%2.5%2.4%2%
Above target but flat to easing - underlying inflation did not justify this hike on its own
HICP ServicesMonthly
OlderPrevLatestTrendTarget
3.2%3.3%3.0%
Easing and at the lowest of the three prints - the domestic-pressure gauge is improving, not deteriorating
Unemployment RateMonthly
OlderPrevLatestTrendTarget
6.3%6.3%6.4%
Still near record lows but off the floor for the first time - marginal loosening, not slack
Negotiated WagesQuarterly
OlderPrevLatestTrendTarget
2.95%2.46%2.44%
Consistent with 2% target - no second-round effects visible in the wage data the ECB itself prioritises
GDP Growth YoYQuarterly
OlderPrevLatestTrendTarget
1.32%0.79%1.0%
Recovering but still below trend - resilient relative to expectations rather than strong in absolute terms
Capacity UtilisationQuarterly
OlderPrevLatestTrendTarget
78.18%77.63%78.52%~81% long-run avg
Well below the long-run average - no capacity constraint anywhere in the data
M3 Money SupplyMonthly
OlderPrevLatestTrendTarget
3.2%3.29%3.38%
Third consecutive acceleration - the one domestic channel corroborating the hawks' 'rates not yet restrictive' argument

Trend Legend: ↑↑ Accelerating up, ↓↓ Accelerating down, Peaked then fell, Bottomed then rose, →→ Stable

Economic Data Divergence

Divergence Level: MEDIUM

Conditions for Hike (0/6 met, 3 mixed)

Inflation persistent and broad-basedMIXED

Headline 3.3% but core 2.4% and services 3.0%, both down; energy 14.3%, goods 1.2%, food 1.2%

Excess demandNOT MET

GDP YoY 1.32% → 0.79% → 1.0%

Tight labour marketMIXED

Unemployment 6.3% → 6.3% → 6.4%

Capacity constraintsNOT MET

Capacity utilisation 78.18% → 77.63% → 78.52%

Financial conditions looseMIXED

Deposit rate 2.25% → 2.50%; M3 3.2% → 3.29% → 3.38%

Wage/cost pressuresNOT MET

Unit labour costs 3.5% → 2.6%; negotiated wages 2.46% → 2.44%

Conditions for Cut (1/3 met)

Inflation at targetNOT MET

Headline 2.8% → 2.9% → 3.3%; core 2.4% → 2.5% → 2.4%

Labour market slackNOT MET

Unemployment 6.3% → 6.3% → 6.4%

Wage pressures containedMET

Negotiated wages 2.95% → 2.46% → 2.44%; compensation per employee 3.5% → 3.3%; unit labour costs 3.5% → 2.6%

Why Holding

Not applicable - the Council acted rather than held. What it declined to do is guide: Lagarde refused to characterise October in either direction ("We are not taking a view as to which direction we go at our next meeting"), because the variable that decides the next move is the persistence of energy prices and their pass-through into wages and expectations, none of which the ECB can forecast from its own domestic data. The genuinely two-sided October meeting is the deliberate product of that uncertainty.

Data to Watch

  • September flash HICP (1 October) - whether headline keeps accelerating or energy base effects begin to unwind
  • Core and services HICP specifically - a third consecutive easing would sharpen the divergence between the hike and the underlying data
  • Account of the 10 September meeting (due ~8 October) - the July account materially rewrote the July decision; expect the same test of whether a further hike was judged likely
  • Q3 negotiated wages and the autumn wage round - the second-round effects test that remains unmet
  • TTF gas prices and EU storage into winter - the account named gas, not crude, as the live inflation channel
  • Euro area M3 and loan growth - the hawks' 'rates not yet restrictive' evidence, now three months into acceleration
  • Middle East conflict trajectory - a de-escalation would remove the entire basis for the June and September hikes
  • Next Governing Council monetary policy decision: 29 October 2026 (Frankfurt)

Policy Evolution Summary

The ECB resumed in September the tightening it began in June and paused in July, delivering a second unanimous 25bp hike to 2.50% as headline HICP re-accelerated to 3.3% on a 14.3% energy surge. What changed is not the domestic inflation picture - core eased to 2.4%, services to 3.0%, unit labour costs to 2.6% and negotiated wages remain at 2.44% - but the Council's assessment of persistence, reinforced by upgraded 2027-28 core projections and by growth revisions that removed July's reason to wait. The decisive point for the next meeting is that Lagarde explicitly declined to characterise October in either direction, which makes 29 October the first genuinely two-sided meeting of this cycle: the trigger is whether energy passes into wages and expectations, not any sign of excess demand.

Analysis generated: 11 September 2026, 07:13 UTC
FP Markets
Trade EUR/USD with FP Markets
$1.50/lot instant commission discount · details
Open account with discount

European Central Bank — Frequently Asked Questions

What is the current ECB stance?

The European Central Bank is hawkish as of its September 2026 meeting. A second unanimous 25bp hike, to 2.50% effective 16 September, delivered exactly as the 27 August account foreshadowed: "another rate hike would likely be necessary unless the inflation outlook improved significantly". It did not improve - August HICP rose to 3.3% from 2.9% on a 14.3% energy surge - so the Council executed. Lagarde framed the decision bluntly: "The conflict in the Middle East continues to generate inflation pressures, and inflation is set to remain well above target for an extended period." The hawkish read is reinforced by the projection round rather than the spot data: 2027 and 2028 core were revised up to 2.6% and 2.3%, leaving underlying inflation above target at the end of the ECB's own medium-term horizon, while growth upgrades for 2026 and 2027 removed the downside-risk argument that justified July's hold. What keeps this short of maximum conviction is the composition: core eased to 2.4% and services to 3.0% in the same print, so the tightening rests on preventing second-round effects from an imported shock rather than on any domestic inflation impulse.

What rate decision did the ECB make at its latest meeting?

At its September 2026 meeting, the European Central Bank hiked 25bp, setting the policy rate at 2.50%.

Is the ECB hiking, holding, or cutting rates next?

Hike probability: medium. Hold probability: high. Cut probability: none.

Get the next analysis the day it’s published

Pick the central banks you care about. We send when there’s a new policy meeting — no weekly newsletter spam.

Central banks

Double opt-in. Unsubscribe with one click. We don’t share your email.