Is “Revenge Travel” the Culprit?

This week we learned that the Consumer Price Index (CPI) for September grew 3.7% compared to September of last year. This was a bit above the consensus estimate of 3.6%. As we all know by now, though, the Fed and others generally exclude food and energy when evaluating underlying inflation trends. Those categories are excluded because food and energy prices are volatile and are generally determined by factors beyond the Fed’s control. The chart below shows the historical growth in CPI on a year-over-year (YOY) basis, both with and without food and energy prices. You can see that the core CPI, which excludes food and energy, is much less volatile. You will also see that the downward trend in core CPI is still intact, whereas the downward trend in headline CPI reversed in August. 

Of course, economists and the Fed also want to look at how inflation changed on a month-to-month (or sequential) basis, and here’s where things get a little trickier. The gray line in the chart below shows the sequential monthly changes in core CPI since the beginning of 2020. We exclude food and energy for the reasons mentioned above. The gray lines for each month are the sum of the orange bar, which represents the contribution from core services inflation, and the blue bar, which represents the contribution from core goods inflation. You will see that core goods inflation reduced the overall core inflation rate for the fourth month in a row, while the contribution from core services inflation increased for the third month in a row. We know that the Fed has been primarily concerned about inflation rates on the services side, so it makes perfect sense that the bond market sold off (resulting in higher bond yields) after these numbers were released.

Now, let’s see if we can figure out what’s causing services price increases to accelerate. First, let’s strip out the cost of shelter, which makes up about 33% of the total CPI and about 57% of the services side (excluding energy services) of CPI. The cost of housing is distorted in the CPI because the methodology used by the Bureau of Labor Statistics (BLS) to track housing costs is flawed. Specifically, the BLS uses all outstanding leases rather than just newly signed leases. Because the growth rate in the cost of new leases has dropped precipitously over the past couple of years, growth rates in rent for the CPI are currently being overestimated by a wide margin. Taking out shelter costs from the services side of CPI produces a sub-index known as the Super Core CPI, which makes up about 25% of total CPI. The chart below shows the YOY and sequential change in the Super Core, which again includes only services but excludes energy services and housing services. You will see that the YOY growth rates in the Super Core continue to moderate, but the sequential growth rate in Super Core increased for the third month in a row to 0.6%. We need to figure out what’s driving the increase in those services categories.

Again, excluding all goods as well as energy services and housing services from the CPI leaves us with the Super Core CPI, which represents just about 25% of the total CPI. However, this one-quarter of CPI is what the Fed has said it’s most concerned about. The Fed is much less concerned about 1) core goods prices (21% of CPI), which have dropped sequentially for four straight months; 2) food and energy prices (21%), which are too volatile to use as a basis for policy which, in any case, might not be effective; and 3) the cost of housing (33%), which is currently distorted due to methodology. So, that leaves us with the following components of CPI that are holding the most sway over the Fed: 

You will see that Lodging Away from Home and Transportation Services are responsible for over half of the sequential increase in the Super Core CPI. So it certainly would appear that revenge travel is the culprit. However, a more detailed look at the components of those categories shows that two sub-categories – Other Lodging, Including Hotels and Motels and Motor Vehicle Insurance – were responsible for virtually all of the increases in those categories. Therefore, we can say that higher hotel and car insurance rates were responsible for about half of the 0.6% sequential increase in the September Super Core CPI. 

It’s hard to say how bothered the Fed might be about higher car insurance and hotel rates. Personally, I think the Fed will heavily discount those price increases as somewhat transitory. Car insurance premiums, which are probably considered a more non-discretionary expenditure by the Fed, are being driven by increased difficulty sourcing parts as well as probably a shortage of labor. Rates could also be growing because people are driving a lot more than they were under COVID lockdowns. Hotel rates are being driven largely by a continued surge in travel demand following the COVID lockdowns. It seems unlikely either of these trends will continue indefinitely. Suffice it to say, then, that the third consecutive increase in Super Core CPI (on a sequential basis) to 0.6% would have been far more benign if we took out housing, car insurance and hotel rates. 

One final thought on inflation. The uptick in headline inflation on a YOY basis over the past three months, driven mainly by higher energy prices, has led to a reversal in the growth rate of “real,” or inflation-adjusted, wages. Slower growth in inflation-adjusted wages could provide comfort for the Fed because it means consumers will have less disposable income to spend on other things. Less spending means less inflationary pressures. 

Bonds and stocks are selling off today as the CPI data was not what some had hoped for. But a more detailed analysis should provide some comfort to investors and the Fed. We remain invested, but defensive, as this historic normalization of monetary (and fiscal) policy runs its course.