Before getting into the minutia of my inferences from the informed consensus tracking of the Core PCE Price Index – which looks better post the PPI – I want to make a comment about seasonality, which is much in the news right now. It is widely believed that there is residual seasonality in the CPI, PPI and thus the PCE Price Index, which tends to make January look quite strong. But that was not a reason to dismiss the beat in the CPI yesterday, because the seasonal issue was known and there was no major surprise in the revision of the seasonals. What is much more interesting is that the details from today’s PPI all cut in the benign direction in terms of their implications for the PCE tracking. (h/t Employ America.) So, the core PCE is now looking to be up 27 ½ basis points, please forgive the false precision. And this, despite the notion that the seasonals will lift January on the month.
Something qualitatively resembling this result was what I had in mind a week ago when I mentioned that the odds were heavily stacked in favor of the 12-month rates of the Core PCE Price Index, my preferred single best measure, and the services super-core version of that all falling meaningfully. As of yesterday, it looked like they were on track to fall, just not meaningfully. But now, they are on track to fall meaningfully. So, my own thinking, which I am careful to update in real time as the news comes in, has roughly made a full circle.
This month, the 12-month inflation rates are going to fall. And according to the informed bean counters I follow, the 12-month rates can be estimated by just tacking on the consensus estimate for the January gain to the old data through December. And this, despite the fact we will be working with new seasonals. This implies pretty strongly that the short-run growth rates should also be close, for example the 3- and 6-month rates. But in fairness, that result does not follow as a mere matter of pure arithmetic. It is possible that October gets revised one way and September gets revised exactly the opposite direction, which would mean the simple extrapolation for the 12-month rate would work (which it will), while that for the 3-month rate would not. But this is a very minor point.
Ok, with that as context. My regular slicing and dicing of the informed consensus, as I read it so far (we get import prices tomorrow), is in the table above, which this month includes actual for December, so you can see precisely how the various inflation rates will evolve.
Compare the January with the December 12-month rates to keep it simple
Source: BEA, FH calculations and inferences
Data are actual to December and estimates for January.
And the chart below shows a history of the core PCE Price Index and my single-best measure through January. Keep in mind that the single best measure strips out some unhelpful sources of noise, but at the cost of introducing a (hopefully stable) downward bias of about 20 bps (ar). So, think of the 12-month rate there, of 1.8%, as implying underlying inflation, according to this measure, of about 2%. Keep in mind also that the single best measure just happens to spot all the benefit of the doubt to the doves in the current context. So, while this is the single best measure, it is not inconsistent with the notion that underlying inflation is still a bit high, just not very.
12-month rates on track to take a new leg lower, as expected — a week ago.
Source: BEA, FH calculations and inferences
Data are actual to December and estimates for January.
Separately, and distantly secondarily, the sectoral mix within the single best measure looks benign. For any given rate of inflation in the single best measure, it is probably better that more come from goods and less from services, because the latter is stickier and more closely linked to the domestic business cycle. Trump raising tariffs will not help on the goods side, but that is a separate – and important – discussion. Right here, I am just describing the current situation. And it is possible that anticipation of tariffs might have artificially lifted trailing goods price inflation anyway, which would reinforce my point about how the mix is at the margin benign.
This mix is better than the opposite would be
Source: BEA, FH calculations and inferences
Data are actual to December and estimates for January.
