Key takeaways
- Euro area HICP inflation rose from 1.7% in January 2026 to 3.2% in May, eased to 2.8% in June, and ticked back to 2.9% on the July flash estimate of 31 July 2026.
- Energy did essentially all of it: energy's contribution went from -0.30pp in February to +0.98pp in May, or 128bp of the 130bp move in the headline.
- Services contributed +1.54pp in February and +1.61pp in May — a 7bp change. Core pressure did not drive the re-acceleration.
- The ECB cut its deposit rate to 2.00% effective 11 June 2025, held it for a year, then raised it to 2.25% effective 17 June 2026 — a hike, and not the 4.00% some accounts still carry.
- With the deposit rate at 2.25% on 31 July 2026 and July HICP at 2.9% on the flash, the real policy rate is about -0.65%.
The 2026 path went up, not down
Euro area annual HICP inflation, on Eurostat's final data: 1.7% in January 2026, 1.9% in February, 2.6% in March, 3.0% in April, 3.2% in May. June came in at 2.8%, confirmed by the full release of 17 July. The July flash, published 31 July 2026, is 2.9%. The chart plots that path. It is a 130bp re-acceleration in three months followed by a partial retreat — the opposite of the disinflation story that dominated euro area commentary at the start of the year.
One detail on the March print is worth keeping. The flash estimate published on 31 March said 2.5%. The final data published on 16 April said 2.6%. A 10bp revision is small, but it is the right order of magnitude for the confidence a flash number deserves — and it is why June's flash was worth waiting on. Eurostat's full June data, published on 17 July 2026, confirmed 2.8% unrevised.
Energy did 128 of the 130 basis points
Decomposing the move is where the story turns. Eurostat publishes each component's contribution in percentage points to the headline rate. Energy contributed -0.30pp in February, then +0.48pp in March, +0.99pp in April and +0.98pp in May. That is a swing of 128bp in energy's contribution across three months.
The headline moved from 1.9% to 3.2% over the same span: 130bp. Services contributed +1.54pp in February and +1.61pp in May — a change of 7bp. Food, alcohol and tobacco went from +0.48pp to +0.36pp. Non-energy industrial goods went from +0.17pp to +0.23pp.
Energy accounts for essentially the entire re-acceleration. This was not a broadening of price pressure; it was a single component moving from a drag to a large positive contribution, while the domestically generated part of the index barely moved.
The ECB hiked, and the rate was never 4.00%
The policy record is unambiguous. The deposit facility rate reached 2.00% with effect from 11 June 2025, ending a cutting cycle that ran 3.00% (December 2024), 2.75%, 2.50%, 2.25%, 2.00%. It stayed at 2.00% for a year. Then on 11 June 2026 the Governing Council raised all three key rates 25 basis points, with effect from 17 June 2026: deposit facility 2.25%, main refinancing operations 2.40%, marginal lending 2.65%.
The stated rationale was direct: "The war in the Middle East is generating inflation pressures." June Eurosystem staff projections: headline inflation 3.0% in 2026, 2.3% in 2027, 2.0% in 2028; core 2.5%, 2.5% and 2.2%; GDP growth 0.8%, 1.2% and 1.5%.
A central bank tightening into a shock it cannot reach
Set the two records side by side. The ECB tightened in response to an inflation impulse that its own data attributes almost entirely to imported energy, at a moment when its staff already forecast that impulse decaying to 2.0% by 2028, and while the services component — the part a policy rate can actually reach — moved 7bp.
Then June landed. The headline fell to 2.8% as energy's annual rate decelerated from 10.8% to 8.5% on the final data — the flash had put it at 8.7% — and services from 3.5% to 3.2%. June's contributions: services +1.51pp, energy +0.77pp, food +0.29pp, industrial goods +0.18pp. The shock had begun rolling over in the same month the hike took effect. Whether that makes the June decision prudent — anchoring expectations before a second-round effect forms — or procyclical is the live question, and the answer is in the services number, not the headline.
The carry arithmetic is stark either way. At a deposit rate of 2.25% against May's realised HICP of 3.2%, the real policy rate is about -0.95%. Against the ECB's own 3.0% projection for 2026, it is about -0.75% for the year. A 25bp hike does not make policy restrictive from there; it makes it marginally less accommodative in nominal terms while staying negative in real ones. Against the July flash of 2.9% it is about -0.65%, and the deposit rate was still 2.25% on 31 July 2026. Euro cash held at that rate has lost purchasing power in every month since March.
What would falsify this
The reading here is that 2026's euro area inflation is an energy shock and not a demand shock. It breaks on one observation: services inflation accelerating above 3.5% while energy continues to decelerate. That would mean second-round effects had formed and the ECB moved early for good reason. Two prints have since arrived and neither triggered it: services was 3.2% in June and 3.3% on the July flash, below the 3.5% threshold in both, with energy re-accelerating rather than decelerating. National dispersion also limits what any aggregate can say: in May, Sweden printed 1.1% and Romania 9.7%. A single policy rate across that range is a blunt instrument by construction, and the aggregate HICP is a weighted average no member state actually experiences.
The July flash of 31 July 2026 has answered the question this piece was built to ask, in the least convenient way. Headline inflation rose to 2.9%, energy re-accelerated to 10.0% from 8.5%, and services went to 3.3% from 3.2%. The decomposition holds — energy is still doing the work — but the print does not settle whether the June hike was prudent or procyclical, because energy turned back up before services had a chance to answer. The deposit rate's next move is a question about growth at 0.8% as much as about prices — and that growth number, and the energy shock behind the whole episode, are now the same problem, and they are visible in front-end rate pricing on both sides of the Atlantic.