Today we’ll continue the exploration of US inflation that we started 2 weeks ago with a discussion of Producer Price Inflation. What does PPI actually measure? Does it help us predict when consumer price inflation might lift off? Inquiring minds, as the old advertisements use to say, want to know.
Three points on this topic:
#1: Here are the component weights for the Producer Price Index (Final Demand) as of December 2020. It’s worth noting that the Bureau of Labor Statistics recently announced it is adjusting some of these weights in the upcoming February 17th release of January 2021’s data. We don’t expect these to materially change the calculations, however.
- Consumer services/Private capital investment: 36.1 pct
- Goods ex food and energy (consumer goods, gov’t purchases, exports): 21.6 pct
- Trade of finished goods (gov’t, consumer, export): 20.3 pct
- Food: 5.7 percent
- Energy: 5.3 pct
- Transportation and warehousing: 4.7 pct
- Government purchased services: 3.0 pct
- Construction: 1.7 pct
- Services for export: 1.4 pct
Takeaway (1): in its current manifestation the PPI is mostly a measure of services, rather than physical goods, inflation. The percentages are basically 33 percent goods, 66 percent services, and 1-2 percent construction.
Takeaway (2): US housing inflation is entirely absent from the PPI but is a critical component of the Consumer Price Index. Recall that “Shelter” is 33.3 percent of headline CPI and 42 percent of core (ex food and energy) CPI.
#2: Goods PPI inflation should be a better “tell” of future core CPI inflation since services tend to scale more easily during periods of economic expansion, so let’s look at PPI goods ex-food and energy inflation and compare it to core (also ex-energy and food) CPI inflation.
Here is that data in chart form back to late 1974 (earliest PPI data available), with goods PPI in red, core CPI in black, and a highlight box showing the December 2020 readings for each.
Two things pop out at us from this graph:
- PPI goods inflation DID NOT predict the growing inflation problem during the 1970s. The red line above actually peaked in February 1975 (far left-hand side) and the follow-on highs in the mid 1980s was lower. CPI inflation (black line) became more of a problem as the decade progressed, as evidenced by its June 1980 high of 13.6 percent versus its February 1975 high of 11.9 pct.
- PPI goods inflation did, however, do a serviceable job of predicting the trend to lower core CPI inflation in the 1980s and 1990s. The red line (PPI) in the chart above was perennially below the black line (CPI) over this period. One could read that as “disinflation in the pipeline”, and it does seem to have filtered through to consumer prices.
Takeaway: note that since 2010 (rightmost quarter of the chart) goods PPI and core CPI have converged, signaling that PPI is no longer leading CPI lower but is in fact running alongside of it.
#3: To evaluate where US inflation is going and what it means for capital markets we need to pull in a market-based forecast. The following chart overlays the inflation expectations built into 5-year TIPS (Treasury Inflation Protected Securities, in green) on the 2003 – present portion of the core PPI/CPI chart we’ve been looking at:
Two thoughts here:
- Aside from the two panic lows in 2008 and 2020 (large downdrafts in the green line), 5-year TIPS have done a reasonably good job of forecasting the stable 2-percent-ish inflation environment we’ve had over the last decade in both PPI and CPI.
- The most recent move higher for 5-year inflation expectations (2.18 percent, the highest since 2013) is therefore significant.
Takeaway: 5-year TIPS are saying that US inflation is heading higher in the near term. Given the close tie over the last decade between core PPI and CPI, we’ll likely see this trend in both datasets as the year develops. Importantly, TIPS are NOT saying rampant inflation is just around the bend. The 2.2 percent forecast imbedded in those bond prices is simply a validation of the idea that the US will see a reasonable and lasting economic recovery in the years ahead.
Final thought (1): the most important takeaway from this analysis is that PPI isn’t as much “inflation in the pipeline” as simply a different take on macro inflation pressures. Headline PPI actually measures something very different from CPI. And yet, as we’ve seen, the 2 track closely together and several decades ago PPI did have the early call on disinflation. Going forward, we’re expecting to see both rise at roughly the same rate.
Final thought (2): in every cycle we’ve covered since the early 1990s some economist or market watcher has tried to make the case that PPI inflation will outstrip CPI and this will cause margin pressures on corporate earnings. We’re sure that we’ll hear this again soon, but it will be as incorrect now as in all prior iterations. Most of the US economy is service-based (as is the PPI), and these businesses can scale more easily (increase revenues without increasing costs) than manufacturing concerns. On top of that, the 5-year TIPS inflation expectations are already telling us that US companies have pricing power – that’s what CPI inflation actually measures. So, rest easy on this issue. We are.