A touch softer
The preliminary estimate of the Harmonised Index of Consumer Prices (HICP) came in a touch softer than our expectations. In our preview, we had flagged downside risks around the consensus (indeed, our forecast was already below consensus). The core HICP reading came in slightly below expectations, driven by Italy (likely a reversal of what occurred last month). Headline HICP was also somewhat softer than we had anticipated, although only by a few basis points.
Based on today’s data, we see no reason to revise our bottom-up forecast. For core, we follow what we did last month and we let the surprise to show up in the level (indeed, in retrospective we could/should have offset partially the Italian surprise, but.. that’s always the case ex-post, isn’t it?). At this stage, we expect headline HICP year-over-year to peak at 3.1 percent in June (2.7% average of year), while core HICP is projected to move roughly sideways at around 2.3-ish percent in the coming months (2.3% average of the year).
Our model-based medium-term projections are little changed. Crucially, both our bottom-up forecast and our model-based forecasts remain closer to the ECB baseline than the “adverse” scenario in this moment.
About the ECB: everyone assumes that the ECB will hike. We understand the point, and the reaction function. However, we ask a simple question: if the current forecast is only a few tenths higher than the March baseline (with little or no signs of second round effects), should the ECB be in a hurry to hike in April? To us, the logic answer is “no”, unless you mark-to-market (but ehi.. that’s the sell-side).
As always, we will await the final data release in order to conduct a more detailed and comprehensive analysis of the underlying distribution.
Charts packages for the EA and for the 4 largest economies are here and here, respectively. (Note: the charts for Italy in the package are not updated, as our dataprovider (Macrobond) is experiencing issues with the source (ISTAT)).
Details
Core HICP is running around 2.0%.
We estimate that core HICP prices grew 22bps MoM (seasonally adjusted) in March, bringing the 3m/3m annualized rate to 2.3%, close to the YoY.
Looking forward, the most probable scenario is that the YoY of core HICP will move sideways.
Figure 1. Core HICP NSA unchained index by year (index normalized to 1 on each New Year’s Eve).
Note: the figure shows the evolution of core HICP by year, normalizing the index at 1 on New Year’s Eve.
Acquired inflation at 1.2-1.7%. With the March data, we estimate that acquired inflation for 2026 is 1.2% for core HICP and 1.7% for headline HICP.
This implies that the risks around the ECB/NCBs forecast are balanced for 2026. Our models are flagging some small upside risks for both core and headline HICP this year and going forward. In any case, the models remain well below the “adverse” scenario of the ECB.
(For a technical explanation of “acquired inflation” and the “carryover effect,” see here and here).
Table 1. Forecast comparison, carryover effect, and acquired inflation.
Medium-Term Model-Based Forecast Little Changed.
Today’s data had no material implication for the model-based forecast, as our Q1 nowcast is little changed.
The model projects core HICP (average YoY) at: 2.3% in 2026, 2.3% in 2027, and 2.2% in 2028.
The model projects headline HICP at 2.7% in 2026, 2.2% in 2027, and 2.0% in 2028.
Figure 2. Core HICP: YoY forecast of our “main” Phillips curve model.
Note: the figures show the YoY forecast of our “main” Phillips curve model for core HICP price inflation. The confidence intervals are calculated with quantile regressions.
Figure 3. Headline HICP: YoY forecast of our “main” Phillips curve model.
Note: the figures show the YoY forecast of our “main” Phillips curve model for headline HICP price inflation. The confidence intervals are calculated with quantile regressions.