What if core PCE prices will expand 0.2% (MoM) each month this quarter?

Keep in mind

The “main” Phillips curve model is expected to revise up significantly the medium-term forecast (2022-2024) should Q3 be similar to last quarter (Q2). On the other hand, assuming core PCE prices will growth at 3.0 percent (annual rate) in Q3 delivers a forecast almost identical to the one at the time of the June FOMC. The model delivers a significantly lower forecast (converging back to target in 2024) only assuming that core PCE prices expand at 2.0 percent (annual rate) in Q3. Given the low probability of this last scenario, we believe the Fed staff will continue to convey a hawkish message to the FOMC. Consequently, we do not expect the FOMC to turn dovish over the summer.

The current forecast

The current “main” Phillips curve model forecast is shown below. The model expects core PCE prices to expand 4.0 percent in 2022 (Q4/Q4), 2.9 percent in 2023 (Q4/Q4), and 2.5 percent in 2024 (Q4/Q4). The current forecast is nearly identical to the one at the time of the June FOMC and very similar to the June SEP. The reason is that at the time of the June FOMC we were already assuming an increase of the unemployment rate in the forecasting horizon, and a deceleration of core import prices in the second half of 2022 given the appreciation of the Broad dollar. (Note: The “main” Phillips curve model is currently run assuming that core PCE prices will expand 4.2 percent (annual rate) in Q2. Under this assumption, the model projects core PCE prices to grow 3.3 percent (annual rate) in Q3.)

Main model forecast, core PCE prices - latest run (2022:Q2 in sample)

A. C.I. calculated from parameters distributions

Note: the confidence intervals (C.I.) are estimated using 100,000 draws from the estimated parameters distributions to simulate the path of core PCE price inflation going forward.

First quarter of forecast: 2022:Q3.

B. C.I. calculated using quasi-out-of-sample methods

Note: the confidence intervals (C.I.) are estimated using quasi-out-of-sample methods (estimate the model over a sub-sample, forecast, and calculate recursevely the root mean squared forecast errors N quarters ahead).

First quarter of forecast: 2022:Q3.

Alternative scenarios for Q3 and model forecasts

What if the “main” model is wrong in its Q3 forecast? What if Q3 will be stronger? How weak(er) does Q3 need to be for the model to revise significantly its medium-term forecast?

The table below shows the model forecasts under three different assumptions for Q3 (in sample). In the first  scenario (line 3 of the table) we assume that core PCE prices will growth at 4.0 percent (annual rate) in Q3, similar to last quarter (Q2). In the second scenario (line 4 of the table), we assume that core PCE prices will slowdown to 3.0 percent (annual rate) in Q3. Finally, in the third scenario (line 5 of the table) we assume that core PCE prices will expand at an annual rate of 2.0 percent in Q3. For each scenario, we report the Q4/Q4 model forecast, as well as the QoQ in Q4 of each year.

Table – “main” model forecasts: current and under different scenarios for 2022:Q3.

There are three main takeaways:

First, should core PCE prices continue to expand at about 30bps per month in Q3, the model will revise up its medium-term forecast in each year (line 3 of the table). Assuming 4.0 percent (annual rate) in Q3, the model delivers a forecast that remains around 3 percent in 2024. The reason for the upward revision is simple: the model currenlty expects a deceleration in Q3 to 3.3 percent; therefore, any upward surprise of the incoming data would translate in a higher forecast.

Second, should prices moderate at annual rate of 3.0 percent in Q3 the model would deliver a forecast just a touch lower compared to the current (line 4 of the table). As mentioned, the model is already forecasting a deceleration in Q3 stopping the sample in Q2. Therefore, per se, a deceleration in monthly prints in Q3 is not enough for the model to revise its forecast, unless the deceleration is bigger than expected. (Note: because of the quarterly arithmetic, a 3.0 percent annual rate in Q3 implies (around) 18bps MoM monthly average in Jul-Sep. In other words, assuming 18bps MoM in Jul-Sep results in a QoQ well above target even if the monthly prints at annual rate would be close to target. In this respect, there is now a potential “tension” in Q3 between the model-based forecast and the signals from the monthly prints, should they moderate. Having said so, the chances to get a sequence of low monthly prints given the monthly distributions appear low at the moment).

Finally, should prices expand at 2.0 percent (annual rate) in Q3 the model would revise its forecast significantly (line 5 of the table). In this case, the model forecast would be around target already next year: 2.2 percent in 2023 and 2.1 percent in 2024.

Implications for the Fed staff

In our view, the Fed staff (and SEP) forecast for the current year is based on a mix of incoming data and estimated fundamentals (slack, underlying inflation, supply-side shocks). On the other hand, the forecast for the next years is based only on the estimated fundamentals. The model-based forecasts are used to assess the risks around the judgmental forecast. Given the evidence of the table above, in our view the Fed staff is likely to continue getting hawkish signals from the model(s), even if the monthly prints moderates in the next 3-4 months. Nevertheless, the probability of having consecutive low readings in core CPI/PCE space appears relatively low at the moment because the distributions of monthly changes are centered at a higher mean/median.

For this reason, in our view the Fed staff and most FOMC participants are not likely to change tone over the summer even at the cost of overshooting (“unconditional” means unconditional). Unless the MoM of core CPI/PCE drops to 0.1 (which again, is not our baseline at the moment), the first moment in which the “main” model is likely to revise down significantly its forecast is in October/November when Q4 will be in sample.

Contact

Want something more tailored?

We provide tailored consulting on ad-hoc projects.