Be careful, very careful. It is January. We expect the NSA level of core and headline HICP at 116.028 and 123.701, respectively. Our forecast for core HICP is consistent with a MoM (saar) around 2-2½ percent, in line with the signals of the distributions. Conditional on our forecast, the YoY on NSA data is expected at 3.25% for core HICP and at 2.85% for headline HICP. Having said that, as we show below, January is by far the most volatile and hard to predict month of the year. Therefore, we invite the reader to be extra careful this month.
Medium-term models below the ECB/NCBs staff forecast (and below 2% for the first time). Conditional on our forecast and a Q1 nowcast of 1.5% (QoQ saar), the medium-term models revised down their forecast. This happened not only because the current quarter is softer than the model expected but also because the seasonal adjustment revised down Q4. The forecast is now below the latest ECB/NCBs staff forecast, signaling downside risks. For the first time, the forecast of one of the models is below 2% by the end of the medium-term.
Our forecast
Be careful, it is January. We expect the NSA level of core HICP and headline HICP at 106.028 and 123.701, respectively in January. Our forecast is based on the assumption that the NSA level of NEIGs prints at 112.390 and services prints 118.150 (the NSA “unchained” level by year of NEIGs and core services can be seen here and here, respectively). Our forecast corresponds to a NSA MoM growth rate of -95bps for core HICP and -28bps for headline HICP. In SA terms, we expect core HICP to expand at around 20bps MoM. The risks around our forecast are well balanced. Our forecast implies the YoY of core HICP and headline HICP at 3.25% and 2.85%, respectively in January. In any case, as usual, we do not put much weight on the sectoral readings, and we will wait for the final distributions.
Note: The “unchained” index of core HICP is shown in Figure 1 (for a discussion about “unchained” HICP see here and here). For the record: the YoY of the unchained core HICP index is expected at 3.0% in January. A chart comparing the YoY of chained (published) vs unchained core HICP is here. The evidence of the unchained index suggests that the YoY of the chained (published) index has room to fall further going ahead.
An important note: please, be extra careful this month. Figure 0 shows the standard deviation of MoM (NSA) core HICP distribution by month. The main message is that January is, by far, the most volatile and unpredictable month (this effect is sometimes called “the January re-pricing effect”). For this reason, we invite the reader to be extra careful, especially because we are coming out of a high inflation period and, given the evidence of Figure 0, it would not be surprising to have a bizarre reading (one way or the other).
Figure 0. Standard deviation of core HICP NSA MoM by month of the year.
Note: the figure shows the standard deviation of the MoM (NSA) distribution by month of core HICP from January 2001 till December 2023.
Figure 1. NSA “unchained” core HICP level by year (1 = new year’s eve)
Implications for the “main” model
Implications for the medium-term model-based forecast of core HICP price inflation. Conditional on our MoM forecast and a Q1 nowcast at 1.5% QoQ saar, the models forecasts are revised down compared to the previous run. The downward revision is driven by two factors: (i) the Q1 nowcast is lower than the model expected for the current quarter, and (ii) by putting Q1 in sample the filters have revised down Q4.
The big news of today’s update is that the models revised down their forecast (and in the end, it could be even lower by the end of Q1, although the seasonal adjustment is introducing a degree of uncertainty here). For the first time, the model using the output gap delivers a forecast below 2% by the end of the medium-term.
Using the unemployment rate as measure of “slack”, the forecast is at 2.3% (average YoY) in 2024, 2.3% in 2025, and 2.3% in 2026. Using the output gap (right panel in Figure 5), the model delivers a more dovish forecast: 2.1% in 2024, 1.9% in 2025, and 1.9% in 2026. Please, note that the risks (intervals of confidence) are skewed to the downside. The average of these forecasts is now below the latest ECB/NCBs staff forecasts.
Figure 2. Model-based medium-term forecast of core HICP (YoY)
Using Urate as a measure of “slack”
Using outputp gap as a measure of “slack”
Conclusion
Cut in March on the table. We are still prudent, especially because everything could change with a very strong reading (which this month we cannot rule out). Having said that, the medium-term models suggest that the risks are now skewed to the downside around the ECB/NCBs staff forecast. In fact, shall we get a reading close to our forecast, a cut in March can be on the table.