A Game Changer For the Models
A game changer. The November flash core HICP figure came below our expectations (NSA level 116.62 vs 117.057 expected by us). November is not a downward surprise like the previous one because it implies a doward revision to the medium-term forecast (next paragraph). While we do not have the details, in reaction to today’s data we have lowered our forecast for December (we now expect the YoY of core HICP at 3.4% in December). Prima faciae, today’s data is so good that it is probably “too good to be true”. In fact, the November NSA MoM is similar to the pre-Covid period and it is compatible with core prices growing below target at the margin. Right now, we are very careful because most of the re-pricing happens at the beginning of the year (see Figure 1 below); therefore, we do not expect to come to a conclusion before (at least) February-March. But today’s data are certaintly changing something.
There are relevant implications for the ECB. Today’s data imply downward revisions to the medium-term model forecasts. The model has revised down its forecast by about 2-3 tenths in each year. The news is very relevant because for the first time, the model forecast is below the latest ECB/NCBs staff forecast (not only in 2023 but also in 2024 and 2025 – see below). Again, this is a game changer because it opens the possibility of downward revisions to the staff forecast even without downward revisions to growth (which are likely).
A PDF containing all relevant charts can be downloaded here.
Details
Too good to be true? We estimate that in November core HICP prices grew 11bps MoM (sa) or 1.4% at annual rate, softer than the previous months (see here and Figure 1). This brings the 3m/3m (ar) to 2.9% (Figure 2). The 3m/3m (ar) is well below the YoY, signaling that the annual variation should moderate further in the coming months. Acquired core HICP inflation for 2023 is 5.0%. Figure 1 now suggests that in the last few months, the evolution of core prices is in line with the ECB target (or possibly even below target). To us, it remains a bit too early to conclude that core will print consistently below 2% going forward (indeed, most of the re-pricing in Figure 1 happens early in the year); but November is certainly a game changer because the series seems to have re-gained the pre-Covid seasonal pattern.
(For the record, we now expect the YoY of core HICP at 3.4% in December 2023. Conditional on this forecast, the carryover for 2024 is 1.1%-1.3% (NSA and SA, respectively))
Figure 1. Core HICP NSA unchained index by year.
Note: the figure shows the evolution of core HICP by year, normalizing the index at 1 on new year’s eve. The shadowed portion of the blue line in 2023 indicates our forecast.
Figure 1. Core HICP metrics.
Note: the figure shows the metrics of core HICP. All figures are seasonally adjusted. “ar”” stands for “annual rate”. The 3m/3m and the 6m/6m are chained (that is, using the US BEA method).
Significant news: medium-term model-based forecast is revised down. Given today’s miss, we have lowered our forecast for December. We are now working under the assumption that core HICP will expand at 2.4% QoQ (saar) in Q4. All other assumptions about the exogenous variables are unchanged.
Conditional on our nowcast for the current quarter, the model forecast using the output gap as a measure of “slack” (Figure 3 right panel below) is at (average of 4 quarters) 5.0% in 2023, 2.5% in 2024, 1.9% in 2025, and 2.0% in 2026. Using the unemployment rate (Figure 3 left panel) the model forecasts 5.0% in 2023, 2.7% in 2024, 2.5% in 2025, and 2.6% in 2026. These figures are about 2-3 tenths lower than before the release (that is: it is a large downward revision for a single HICP print).
Today’s revisions are very relevant: for the first time, the model (using the output gap) is delivering a forecast below the ECB/NCBs staff forecast in 2024 (2.5% vs 2.9%) and 2025 (1.9% vs 2.2%). In our view and estimates, this opens the gates to a downward revision of the NCBs staff forecast in December throughout the entire medium-term horizon.
Figure 3. Core HICP: YoY forecast of our “main” Phillips curve model.
Using Urate as a measure of “slack”
Using outputp gap as a measure of “slack”
Note: the figures show the YoY forecast of our “main” Phillips curve model for core HICP price inflation. The confidence intervals are calculated from the estimated parameters distribution. Last quarter in-sample is 2023:Q4.
Implications for the ECB staff and monetary policy
We continue to expect a prudent ECB. Today is a good day for the ECB/NCBs staff and the Governing Council. As shown above, the downward revision of the models forecast is sizable and it is hard to think this will not be reflected in the December forecasts. Having said that, given the status of the labor market, it is also hard to imagine a forecast below target, at least in the December projection unless the staff will fold-in a recession, which is unlikely in our view. For this reason, we expect a lower forecast (similar to the model using the output gap in Figure 3) but we expect the Governing Council message to remain very prudent (“high rates”). It is still too early to talk about cutting rates.