Some indicators (i.e. Manheim Index) point to possible downside risks in August for consumers prices. In this note we answer the question “what can be the Fed staff reaction should we get a soft (core) reading in August”?
Keep in mind
Keep in mind. In case of a soft (core) reading in August the Fed staff might revise down its 2022 forecast to 3.8%. However, our research suggests that the medium-term forecast (model-based) and the estimation of trend inflation will be affected only marginally. For this reason, we continue to believe that a real Fed pivot will require a sequence of soft readings in Fall.
The data
In July we estimate that core PCE prices have expanded 14bps (31bps in core CPI space). As for August, some indicators (i.e. Manheim Index) point to further softness. In our estimates core CPI (PCE) could expand 0.2% (0.1%) MoM in August.
How would the Fed staff react to a second soft core reading?
#1 The judgmental near-term forecast
In our view and experience, the Fed staff near-term forecast (4-6 months ahead) is a judgmental forecast based on ad-hoc models, on the statistical properties of the incoming data, and on the narrative of the monthly prints. For this reason, it is hard to say anything in advance because the Fed staff reaction strictly depends on the data. However, the general rules to update the near-term forecast are: (i) take signal from persistent items (i.e. core services, especially housing), (ii) discount/offset weakness in volatile categories (i.e. core goods), (iii) discount/offset weakness coming from idiosyncratic shocks affecting a limited number of items and, (iv) take signal from shocks common across items (i.e. the CI-C model).
Therefore, unless core services will slow significantly in August, the Fed staff in our view will take only little signal from the incoming data and possible revisions to the near-term forecast should be more cosmetic than substantial. Having said so, we would not be surprised to see the Fed staff revising down their near-term forecast to 3.8% in December 2022, just a touch lower than the model-based forecast (see point #3).
#2 The judgmental medium-term forecast
In our view and experience, the Fed staff medium-term forecast (2-3 years ahead) is a judgmental forecast based on a calibrated augmented Phillips curve model. We defer the discussion of the details of such calibration to private talks (please, refer to the medium-term “decomposition” in our “Pre-FOMC Meeting Package”). For what matters in this note, according to the July FOMC minutes the Fed staff was expecting core PCE price inflation at 4.0 percent in 2022, 2.6 percent in 2023, and 2.0 percent in 2024. In our view, a low August reading of core inflation would translate in a downward revision to the 2022 forecast (to around 3.8%) but not to the 2023 or 2024 forecast. The reason is that the medium-term decomposition is currently suggesting a large positive residual for 2022 (put it differently: the model does not understand part of the strength of the data). Therefore, a low reading in August would marginally alleviate the issue of a large residual in 2022 but would not influence the 2023 and 2024 forecast.
#3 The risks around the medium-term forecast
As usual, we assess the risks around the Fed staff forecast using our “main” Phillips curve model. The table below shows the current forecast and the one including a low (0.1%) core PCE reading in August. In both cases we include 2022:Q3 in the sample but in the “current forecast” we assume that core PCE prices will expand 3.6% (a.r.) in Q3 while in the “low august” scenario we assume 3.3% (a.r.) consistent with a 0.1% core PCE reading in August. Table 1 and figure 1 show the current forecast and the “low August” scenario.
Table 1 – Core PCE price inflation forecast
Note: the table shows the forecast of core PCE price inflation from our “main” Phillips curve model. The “Current forecast” columns show the latest forecast. The “Low August” columns show the forecast assuming a weaker Q3 (3.3% at annual rate instead of 3.6%).
Figure 1 – Core PCE price inflation forecast
Main takeaway
The main takeaway from this exercise is that the impact of a possible low (core) August reading on the model-based forecast is very small. There are two reasons why this happens. First, the downward revision to Q3 triggered by a low reading in August is relatively small. Second, the model estimates a high persistency of the inflationary process (that is, the sum of the autoregressive coefficients is now higher than pre-pandemic: 0.67 vs 0.52). Therefore, the starting point of the forecast is marginally less important and the forecast converges towards pi* more gradually.
(We have run our set of 11 trend models assuming a lower Q3 as we have done with our “main” Phillips curve model. The impact is negligible and the average across models remains at 2.7%).
Conclusion
Our analysis shows that in case of a soft (core) reading in August the Fed staff might revise down its 2022 forecast to around 3.8%. However, our research suggests that the medium-term forecast (model-based) and the estimation of trend inflation will be affected only marginally. For this reason, in our view it will be hard for the Fed staff to deliver a dovish message to the FOMC at the September round. Rather, we continue to believe that a real Fed pivot will require a sequence of soft readings in the Fall, a scenario which seems unlikely at the moment. The Fed might slow the hikes after September but it is hard to imagine a fast shift to cuts in early 2023.