Given the requests, we circulate an updated FRB-US baseline and we show under what conditions the model delivers two cuts in 2024. Compared to the last update, not much has changed. The baseline remains one cut only. The model delivers two cuts assuming either a protracted slowdown or very low core PCE prints. We think Powell is not bluffing, the Chair is taking time because while recent data have been on the soft side, the models are not there yet. Therefore: if the question is “will the model likely show two cuts at the time of the September meeting?”, the answer is “probably not”. If the question is “should/will the Fed cut in September?”, the answer is “according to the model, no unless the FOMC accepts the risk of remaining a bit above target in order to save the labor market”. We suspect Powell has not made up his mind yet. The final decision will be judgmental.
At the end of this note, please find a word of caution about the upcoming CPI. We perceive the risks around our forecast to be skewed to the downside. We remind the reader why.
Updated assumptions
The new baseline is constructed assuming: (i) Q2 GDP growth at 1.8% QoQ saar, (ii) the unemployment rate at 4.0% in Q2, and (iii) core PCE price inflation at 2.8% in Q2 (QoQ saar) and 2.3% in Q3 (QoQ saar).
Scenarios
The model can deliver two cuts in 2024 but only under particular assumptions. The idea is to test under what conditions (lower GDP growth, and/or lower core inflation) the model delivers the second cut. We simulate two scenarios. In the first one (Figure 1), we reach our goal of two cuts in 2024 lowering real GDP growth. In the second scenario (Figure 2), we reach the goal lowering core PCE price inflation. In each chart, the red dashed line is the current baseline, while the solid gray line is the scenario.
The conclusion is the following. Scenario #1 implies real GDP growth at 1% QoQ saar for 4 consecutive quarters. Scenario #2 implies 50bps lower core inflation rate in each quarter (QoQ saar) for 4 quarters compared to the baseline (i.e. core PCE around 1.7% QoQ saar or lower from Q3 onwards). Assuming a combination of Scenario #1 and #2 delivers 2 and a half cuts in 2024.
Figure 1. Scenario #1 – Lower GDP Growth
Figure 2. Scenario #2 – Lower Inflation Scenario
Note: Real GDP growth and core inflation are expressed as YoY. Core inflation is core PCE price inflation. The dashed red line shows the current baseline, while the solid gray line shows the scenario.
Conclusion
This exercise is not meant to provide exact simulations. Instead, it is meant to deliver the following message: any single CPI/PCE/Urate print has a small impact on the model(s), way smaller than perceived and generally priced by the markets. According to our simulations, it is unlikely that the FOMC will have much support from the models to cut rates in September, even if we have favorable data in the next two months. Ultimately, it seems that if the FOMC will cut, it will do it only judgmentally.
Why we argue there are downside risks around our June CPI forecast
In our CPI preview (here), we wrote that we see downside risks. We explain why. When putting together the forecast, we start by using published SA data, we forecast in SA space, and then check the implied NSA level (using NSA data in our models would be way more complicated or virtually impossible). The issue in June is that the NSA level implied by our SA forecast looks very high and could easily turn out to be lower (dashed line in Figure 3). Indeed, if we had to judgmentally adjust our forecast based on the NSA level, our MoM forecast in SA space would be 8bps lower for both, core (15bps instead of 23bps) and headline (1bps instead of 9bps). June is the first month this year in which we noticed such a discrepancy between our SA and NSA forecast. Therefore, we invite the reader to be very careful. In any case, even if we print on the dashed line, the NSA level would continue to be consistent with a 3%+ forecast YoY in December 2024 and, as mentioned above, it would change little for the Fed.
Figure 3. Cumulated NSA level of core CPI by year
