A very good CPI report, above consensus. More important is August’s core PCE. A translation suggests a 0.15% increase in core PCE in August. Very promising, as a 0.15% monthly print is 1.8% annualized. But, of course, it’s just a data point and not an indication of the underlying trend.
The December hike decision, in our view, depends principally on an assessment of underlying core inflation. The statement will, as always, say that the Committee will look at a wide range of indicators, and that is especially the case given recent inflation readings.
Usually, the FOMC gives almost exclusive attention to the 12-month core PCE inflation rate as a measure of underlying inflation. That peaked at 1.9% in February 2017 and was 1.8% to 1.9% from July 2016 to February 2017. That appeared to be the underlying rate. If the FOMC were comfortable with that, then December would be a go, no doubt about it. But that measure has fallen about ½ pp since February, and will likely fall to 1.3% in August. But no one thinks that the underlying rate is actually as low as 1.3%.
For one thing, price level shocks are playing a role, and these effects are transitory, but lingering as long as they are still in the 12-month window. Declines in cell service prices alone took one to two tenths off 12- month core PCE inflation in March. The trend is, therefore, more likely to be closer to 1.6%. That’s the same as the most recent 12-month rate for the Dallas Fed trimmed-mean measure, which is perhaps more reliable when there has been a price-level shock other than in oil or food. We now view the underlying rate as 1.6%, but that rate is less clear now. We also expect that the Committee will pay more attention to shorter-horizon measures of core inflation, specifically from just after the March decline in core prices through October, the last date they will have in hand.1 This report moves in the right direction.
As shown in the table below, the 12-month rate would be 1.4% in October, even assuming two more good reports like this. Still, that would reflect some downward bias because of the price-level shock. But the rate from March to October would be 1.6%. Two more months of inflation reports like this one might give the Committee enough confidence in its inflation forecast—getting to 2% by 2019, if not 2018—and make them comfortable hiking in December.
So a December hike was a close call. Closer after today’s report. I Will stick with no hike, but the probability of a December hike has risen toward 50/50.
1 November CPI will be released on December 13, the same day as the FOMC statement. But that release is very unlikely to affect the FOMC’s decision—given how late in the FOMC’s decision-making process it will arrive and that, in general, one data point provides little additional information about the underlying trend.
Core PCE Inflation