We have updated FRB-US to include Q2 in-sample. This note explains the new baseline (a touch weaker than last run), as well as two more favorable scenarios in which inflation falls more quickly than expected. The takeaway is that the distribution of risks is now well balanced around the latest SEP with some downward risks in the near-term. We explain why it is now hard for us to bet against the Fed (and the ECB) and what it would take to change our assessment.
Including Q2 in-sample
A touch higher unemployment rate, lower core inflation, and slightly lower FF rate. Because the dataset consistent with the June SEP is not public yet, we are still working with the inconsistent dataset (that is, net of the add factors of the Fed staff) of the March SEP. As such, these results should be taken as preliminary and we will update them again once the new dataset will be released. As usual, we have updated the relevant variables (GDP details, core PCE price inflation, unemployment rate, etc..) through Q2 using our nowcast. The inclusion of Q2 resulted in a (small) downward surprise in GDP space as the model forecasted 2.4% (QoQ ar) but we are working with a 2.0% nowcast. As for the unemployment rate, our nowcast (3.6% end of period) is slightly higher than the model forecast. Finally, our nowcast for core PCE price inflation in Q2 is a bit lower than the model expected.
(Two technical notes. 1) As mentioned, the model forecast for Q2 real GDP growth was 2.4% and it has not really changed since we have been monitoring it. At the same time, Blue Chip consensus for Q2 at the end of April was still at 0% (put it simply: FRB-US was way ahead of consensus). 2) In the model, the revisions to GDP growth and the unemployment rate do not have to be consistent with the Okun law at high frequencies (quarter-by-quarter) because the Okun law error can be non-zero at any point in time).
Figure 1. Updated baseline forecast of “inconsistent” FRB-US
Note: Real GDP growth and core inflation are expressed as YoY. Core inflation is core PCE price inflation.
The inclusion of Q2 results in a peak of the FF rate slightly lower (5.8% vs 5.9%) than in our previous update. The downward revision is driven, as mentioned, by an economy which is a less strong than expected (by the model). As for core inflation, the model now expects it at 3.7% in 2023 (Q4/Q4). In this updated baseline, the Fed starts cutting the FF rate in 2024:Q2 when the unemployment rate is forecasted at 4.2% and core PCE price inflation (YoY) is at 3.3%. As a reminder, FRB-US does not have a banking sector; therefore, considering the estimated impact of the credit crunch, we consider the “true” peak of the FF rate being about 30-50bps lower than in Figure 1.
(The reader should notice that in this update FRB-US expects core PCE price inflation (Q4/Q4) below the June SEP. This result is still preliminary given the current dataset. However, this is the first time the model delivers a forecast more favorable than the SEP)
Figure 2 shows the stochastic simulations around the new baseline. Following standard procedures of the Fed staff we draw randomly from historical errors (1970q1 to 2017q4) for 54 variables and 5,000 replications. The simulations continue to show that: (i) calling a recession in this environment is like a flip of a coin, and (ii) conditional on the estimated persistency, the model sees a (very) low probability of getting back to the inflation target by the end of the medium-term.
Figure 2. Stochastic simulations around the baseline.
Note: Real GDP growth and core inflation are expressed as YoY. Core inflation is core PCE price inflation.
More favorable scenarios
Hard for the Fed to cut, unless inflation slows down to target sequentially (which is unlikely for now). We have simulated two scenarios in which core PCE falls more quickly than in the new FRB-US baseline. The reason is that our models currently suggest that the MoM of core CPI in Q3 could be quite favorable. In particular, while we expect core CPI to remain solid in June, we currently expect the following MoM in Q3: July core CPI 19bps, August core CPI 22bps, and September core CPI 25bps (at the same time, we expect core CPI to rebound back in Q4). For this reason, we simulated (“Scenario1”) that core PCE grows at 3% annual rate in both Q3 and Q4, and that (“Scenario2”) it grows at 2% annual rate in the last two quarters of the year. In other words, “Scenario1” is close to what we suspect will happen, while “Scenario2” provides a lower-bound forecast for the FF rate. (Important technical note: in both scenarios we have built-in a shock to core PCE price inflation in Q3 and Q4, but let the forecast unchanged/unrestricted from there onward)
Figure 3 shows the path of the FF rate in these two scenarios. In Figure 3 the blue line is the baseline presented in Figure 1, while the orange lines are the respective scenarios. As mentioned, we have shocked core inflation only in two quarters, not its entire path. Therefore, the evidence in Figure 3 is indicative for the peak of FF rate but not (necessarily) for the level in 2025. Under “Scenario1” the peak of the FF rate is about 15bps lower than in the baseline, while in “Scenario2” the peak is about 30bps lower than in the baseline. Overall, these results seem to suggest that in case core prices were slowing down to a pace of about 3%, the Fed would not materially deviate from the path signaled in the June SEP. A more pronounced slowdown would produce an earlier cut, although such scenario seems unlikely at the moment.
Figure 3. FF rate under more favorable scenarios
Note: the blue line shows the baseline presented in Figure 1. The orange lines show the two more favorable scenarios.
Conclusion
Careful now. In the last 18 months we have constantly bet against the Fed (and even more so against the ECB). The reason is that our models suggested that the central bank and its staff were late and/or wrong. The June meetings both for the Fed and the ECB brought the central bank(s) roughly in line with the models. Not only, but the distribution of risks in the next quarter seems to be skewed to the downside. Disinflating an economy takes time and we are still on a minefield (UK, Canada, Norway, etc..) so we cannot exclude an upward surprise in core inflation space for the US. But at this point it is hard for us to bet against the Fed because it would take a surprise. It can happen, but again at this point we would need another (upward) surprise, given what the models are suggesting and what the Fed is forecasting.