In 2019 market participants have viewed moves in inflation asymmetrically. Phases of decline, although largely noise, they treated as concerning – thinking that perhaps it’s a slide into deflation (an unlikely event historically). By contrast, the recent acceleration has been largely ignored.
Primarily, this divergent perception reflects a skew on the part of FOMC policy makers. They tend to treat downside inflation surprises more seriously than upside surprises in a world where inflation has been slightly below target since it was introduced in 2012 (never mind that it was substantially above the Fed mandate of price stability for about 50 years before that).
Recent Fed statements suggest that the FOMC might well respond to higher inflation by refraining from raising interest rates anew until inflation breaches 2.5%. I think this approach is unduly risky, given the cyclical state of the economy, but it is tenable if the bond market goes along with it. Refraining has the advantage of aligning with the political calendar, since it is unlikely for any inflation acceleration to broach 2.5% until after the election. This leaves the election winner with a potential problem heading into 2021, however.
Inflation is not Unusually Low
Much of the narrative for US inflation in 2019 depends on the notion that it has been unusually low. There is also a widespread perception that inflation has behaved differently from the rate projected by Fed models: given such a low unemployment rate, the inflation rate should surely be higher. But I would caution against such new age thinking.
A quick scan of inflation rates since 1994 does not suggest that the recent trend in inflation has been particularly unusual (Chart 1). As is almost always the case, the core personal consumer expenditure deflator has risen as a slower clip than the core CPI (the gap averages just below 0.5%-point; in September, it was 0.7%-points).
Looking at the last two late-cycle episodes in the post-1994 era, we see a paradox: the late 1990s experienced core inflation that was even lower than it has been in recent quarters. Core inflation was higher during the 2005-6 late-cycle episode, when there was greater slack than either now or in the late 1990s.
Other Measures of Inflation have Picked Up
Other central banks have developed alternative methods of measuring core inflation. Some of these measures have been adapted to the United States by regional Federal Reserve Banks. The Atlanta Fed has developed the ‘sticky’ CPI—a measure of inflation where the components change relatively infrequently (about 2/3 of the CPI and dominated by services). The Cleveland Fed has developed a trimmed-mean measure, which excludes outlier m/m price increases, whatever their source may be. Both measures have accelerated significantly from low readings nine years ago.
Meanwhile, a simple inflation framework used by former Fed Chair Janet Yellen in a 2015 speech has worked well in 2019. This model posits inflation to be driven by its own lagged values, by inflation expectations, by a measure of aggregate slack, and by the relative price of imported goods. This model does a pretty good job of capturing recent swings, including the dip down from a peak in 2018 to a recent trough in 2019 (Chart 2).
