Provocative question: who could have guessed it? Answer: our models. How many economist/forecasters in, say, December 2022 forecasted a 0.5% MoM sa four months out (that is for the April 2023 core CPI reading)? To our knowledge, the answer is zero. But here we are: there is a chance next week we will get a 0.5% MoM sa core CPI, and then a 0.4% MoM sa in May. Who could have predicted it? Our medium-term models, is the answer. The unpleasant corollary is that the same models continue to send uncomfortable signals (see paragraph below) for the future.
(Unrelated note: today, the BLS informed that the Unit Labor Cost in the nonfarm business sector increased 6.3% ar in Q1, in line with the previous 4 quarters. As Jason Furman correctly framed it, the reading “Suggests that thinking there will be a big round of margin compression to bring down inflation without slower wage growth is possible but not probable”. This is exactly what we wrote yesterday when commenting Powell’s Q&As.)
Our forecast
High core, little or no progress on disinflation. We expect headline and core CPI to expand 52bps and 48bps (6.4% and 5.9% at annual rate) in April, respectively. Figure 1 below shows the sectoral breakdown of our MoM forecast for April. We expect core goods and core services to expand 65bps and 40bps, respectively. Conditional on our forecast, the disinflation progresses would be limited, as the MoM of core CPI would be in line with the average of the previous 17 months. As usual, we do not comment on the sectoral breakdown because we give much more importance to the ex-post distribution of price changes. Finally, as mentioned at the beginning of this note we expect core CPI to remain strong (0.4% MoM sa) in May. Figure 2 shows our MoM forecast errors: in the last 6 months, for core CPI the standard error of our forecast is 3bps and the standard deviation is 8bps (with our forecast slightly biased to the upside by 2bps).
Figure 1. MoM CPI forecast – details
Figure 2. Forecast errors
Implications for the “main” model
Implications for the medium-term model-based forecast of core PCE prices can be significant. Because we will include Q2 in-sample for the first time, we will evaluate the exact implications next week. Having said so, should we get a strong reading, the model is likely to revise up (significantly) again the entire path. The Figure below shows the run of the model at the time of our pre-May FOMC meeting package. But again, the new forecast will most likely be (much) higher. We will evaluate the exact implications and how to deal with the model after the CPI report.
Latest forecast of our “main” model for core PCE price inflation (YoY).
Conclusion
It takes time to disinflate. If our forecast is correct, the Fed staff (and the FOMC) will be, once again, surprised upward by the incoming data. To us, the question is the same: if the data remains so strong and the Fed is “data dependent”, how can it cut the FF rate during the summer?