A false positive disinflation signal
The February PCE price report is a “false positive disinflation signal”. The reason is simple. Market-based core PCE prices grew 38bps MoM (or 4.7% ar). The softness in today’s report comes from non-market prices, which managed to contract, to be revised down in January, and which carry no signal for future readings of core PCE itself (see figure 5 below). Our models and distributions confirm this diagnosis. In our experience, the Fed staff will take no signal from non-market prices. Overall, zooming-out signs of disinflation remain limited.
A note on non-market prices: much more unpredictable than pre-Covid. Other-than market based core PCE prices (a.k.a. “non-market” prices) have become much more volatile and less predictable in the last two years. The standard deviation of the MoM of non-market prices increased from 2.7 (2016-2019) to 9.1 (2020 to present). Unfortunately, a large portion of the non-market prices is almost impossible to predict on a monthly basis because the BEA is very vague about its construction. With this note, we simply point out that for a given CPI reading, the MoM of core PCE can easily round up/down 1 tenth (or even 2) for a 1 std dev shock to non-market prices (that is, pretty frequently), with limited possibility to forecast them correctly on a consistent basis.
Evidence from the distributions
Distribution sending unclear signals. This month, most percentiles of the distribution (not shown for brevity, available upon request) moved down, although they are comparable to their December levels. The Kernel of the last three months (Figure 1) is sending now mixed signals: the distribution is trying to regain its pre-Covid mean but the right shoulder of the last three months is thicker than the previous 3 (or 6). Therefore, it is very difficult to take some signal from the distributions now. We need much more convincing evidence to conclude that the real disinflation has begun. The median of the distribution has ticked down this month (Figure 2) but remains well above pre-Covid (again, we do not put too much signal on the movement of the median because we have already seen similar movements in previous months. Overall, in the last 18 months the median has been volatile and moving sideways).
Figure 1. Kernel of PCE excluding food and energy items changes (%, a.r.)
Note: the Figure shows the fitted Kernel (Epanechnikov) distribution of MoM percent changes at annual rate of PCE prices excluding food and energy items.
Figure 2. Median PCE price increase
Note: the Figure shows the median (MoM %, a.r.) of the distribution of PCE prices changes excluding food and energy items (left panel) and the YoY (right panel).
Evidence from our CI-C model
Our CI-C model estimates that net of Covid and idiosyncratic shocks, the strength of the data in February is similar to recent months. Figure 3 shows the decomposition of the MoM of core PCE in the “common” component, the “idiosyncratic” component, and the “Covid” effect. The model estimates that in February the common component increased by 20bps, a bit less than the previous 2 months (24bps and 21bps, respectively), although higher than the readings at the end of last year. The Covid effect is estimated at 7bps, and the idiosyncratic shock is also positive (4bps).
Figure 3. Contributions to MoM changes of PCE excluding food and energy items (CI-C model)
Note: the Figure shows the decomposition of the MoM percent changes of PCE prices excluding food and energy items. The contributions are estimated using our CI-C model, a 2-stage OLS-LASSO regression model. The “Covid” effect is identified with price variations outside the 10th-90th percentiles of each item pre-Covid price change distribution.
Figure 4. Estimated “Common” component: YoY, 3m/3m a.r. and 6m/6m a.r.
Note: the Figure shows the 3m/3m at annual rate (green line), the 6m/6m at annual rate (red line), and the YoY (blue line) of the “common component” estimated using our CI-C model.
Implications for the medium-term forecast of core PCE price inflation
The medium-term forecast of core PCE is little changed. Compared to the time of the March FOMC (here the slides) the model delivers a forecast which is sligthly lower, as today’s data mechanically pushes down the starting point. In any case, the downward revisions are more cosmetic than substantial. The model (Q4/Q4) forecast is: 4.4% in 2023, 4.0% in 2024, and 3.7% in 2025 (Figure below).
Note: the figure shows the latest run of our “main” Phillips curve model. The confidence intervals (C.I.) are estimated using quasi-out-of-sample methods (estimate the model over a sub-sample, forecast, and calculate the root mean squared forecast errors). First quarter of forecast: 2023:Q2.
Implications for the Fed Board staff
Today’s PCE reading carries little good news for the Fed staff. If we put everything together, we are left with little progress on the inflation front for a full year. Figure 5 below delivers the same message by showing the MoM (ar) of core PCE prices market-based (from which the Fed staff takes a lot of signal, in our experience). As Figure 5 shows, net of volatile items (non-market prices), we have gone sideways since the beginning of 2021. In our experience, the Fed staff is well aware of this and should deliver consequently.
Figure 5. MoM of core PCE prices, market-based vs other than market-based.
Note: the Figure shows the MoM at annual rate of core PCE prices market-based (blue solid line) and other than market-based (light blue dashed line).