Progress on Inflation has Stalled Alright...at just about 2%.

Articles have started to accumulate that progress on inflation has "stalled out." See here, here, here and here for just a few examples in recent days and months making these claims. Is this actually the case though? Has our progress stalled out? November's CPI report this morning shows that if anything, we've stalled out alright...right around 2%.

Let's go through November's report. For the month, headline CPI (which includes food and energy prices) rose 0.3% month-over-month (and seasonally adjusted), while core prices (which exclude food and energy) rose by a similar amount. Unrounded, I estimate core prices rose by 0.28% (28 bps). As the table below shows, this puts the year-over-year figures at 2.7% for the aggregate CPI index, and 3.3% for core CPI. As the table below also shows, this is the second straight month where the headline CPI's year-over-year figure has increased sequentially, and the 6th straight month where the core index has risen at 3.2% or 3.3% year-over-year as well. That certainly looks like "stalled" progress at first blush. So why do I think otherwise?

At first blush, looking at the year-over-year figures for the last 3-6 months could easily make you think that progress on inflation has in fact "stalled out." But as we'll show in a moment, this is misleading, and if anything, "progress" on inflation has "stalled" out right where we want it to...around 2%.

As we usually do, let's go through our three "buckets" of inflation: goods, services ex. shelter, and shelter.

Goods was the most noteworthy this month, contributing +8 bps in November to our core increase of +28 bps. Similar to last month, car prices were a big driver, contributing 7 of the 8 bps of the core goods increase. New vehicle prices contributed +3 of the +7 and used vehicle prices contributed +5 of the +7 (note that 3 + 5 doesn't add up to 7 because of rounding...new vehicle prices contributed 2.7 bps and used contributed 4.7, equating to 7.4 bps as a sum, which rounds down to +7). Notably though, both new cars and new trucks contributed roughly 0 of the 2.7 bps from new vehicle contribution, and those are the only sub-categories, so it's not clear what was driving this category (those are the only sub-categories of New Vehicles we get).

Used cars and trucks were again a large contributor this month, except this month we didn't have the -5 bps of contribution from apparel prices to offset the contribution from used car prices. This is what caused a core goods "miss" this month. As the second chart below shows, used car prices may have stopped declining, but it's far from clear if they're going to actually start increasing again. Language matters a lot with inflation: flat inflation, which would be good, does not mean declining prices. These days, if prices aren't falling, often times we're still told that isn't good enough, even though it's quite rare for prices to fall.

Manheim's Used Car Index has shown a slight tick up lately, but it's far from clear whether we'll actually see regular (notable) increases from here. Given the volatility on a monthly basis, we'll have to wait and see.

Excluding cars then, core goods increased +1 bps for the month. Given cars historically don't inflate much (as shown above in the other charts), and given rates continue to remain elevated, it's not clear why we should assume that the kind of increases we saw in November in the CPI would persist going forward. Of course, we'll be watching.

Now let's turn to services ex. shelter, otherwise known as "core services" or "super core." For the month, core services increased +6 bps, which is slightly below the pre-COVID average of +7 bps. This was probably the bright spot of the report this month, especially after the last few months have been a little stronger in this area. Notably, car insurance, which has been a real thorn in the side of core services, was essentially a non-contributor again this month.

"Super core" or "core services ex. shelter" for the month was below the pre-COVID average, and that was inclusive of notably high airline and hotel inflation for the month.

Let's now turn to shelter costs, which have been the bugaboo that we've talked about for months in our CPI reviews. This month, shelter costs added +14 bps to the monthly core total of +28 bps. As the chart below shows, this is seemingly (finally) trending in the right direction, but remains elevated despite private market figures indicating that rents are growing slower than they were pre-COVID (not faster, as the CPI data indicates). See here, here and here for (free) private market data on this.

Shelter costs remain elevated relative to pre-COVID averages, despite private market data indicating rents are growing at below historical levels at the moment, not higher.

Putting all this together, if we normalize for shelter, inflation remains around 2% in November (as it has for a while now). You might be asking, how can we feel so comfortable normalizing for shelter? The answer is two reasons: first, because of the huge discrepancy between multiple private market sources (which we linked to above) and the government data; second, because we know the government's methodology produces a lag. As a reminder here, the government's methodology can be summarized by the example of an apartment building with 12 units, each of which renews their lease in each month of the year (so one apartment renews in January, one in February, and so on). If each renewal is a 3% increase, private market data will indicate that "market" rents for this building are up 3%, beginning in January. But the way the government will calculate the data, it will only use January's 3% increase multiplied by 1/12, and will use the renewals taken from the previous 11 months the year before in the rest of the calculation. This produces a significant lag.

The chart below helps demonstrate what inflation today looks like if you normalize for this shelter lag and if you take averages for core goods and services to adjust for some of the monthly noise (like we saw with vehicles this month). Notice that the bars and lines in the chart are all just above the 2% line as of November. The most recent month's figures using this methodology produce a 2.35% annualized inflation rate using November's report, 2.56% using a 3 month average annualized rate, and 2.2% using a 6 month average annualized rate. Again this is adjusted for the shelter lag. But how do we actually adjust for the shelter lag? To be conservative, I simply assume shelter contributes the +11 bps that it did before COVID. Importantly, this is actually well above what shelter is inflating in the real economy today. So if anything, these estimates of "run-rate" inflation figures are overstated, not understated. Thus, even with an adjusted rent figure that is likely overstated (and thus likely overstates our total estimate of inflation), we're either at, or within a stone's throw of our inflation target. This was just as true in November as it was in recent months.

Thus, while it's easy to conclude that progress on inflation has "stalled out" well above the Fed's target, that's only true if you look at the headline optics, which are meaningfully distorted by shelter. Though you'll likely read plenty more articles today about our "stalled" progress on inflation, you likely won't read many that we're stalling out in a very good place: 2%. Thankfully, the Fed has been smart enough to look through both the noise and the lags (particularly from shelter) and realized that making the proper adjustments, it's time to start cutting interest rates. It probably shouldn't be surprising then that the market odds for a rate cut at next week's Fed meeting went from 87% yesterday to 99% today.

That being said, because of all these misleading articles, even Fed officials can probably only go so far in ignoring the reported figures in each month's CPI report. Let's hope that future months see further downward movement in shelter costs in the CPI so that they don't have to negotiate misleading inflation optics and instead can do what's best for the economy. It's not clear how many more rate cuts this economy will need as time goes on, but one less complicating factor would certainly be helpful.

In case this is the last time you read this blog this year, Merry Christmas and Happy Holidays, and look forward to a lot of new content in 2025!