Dec. '24 CPI Report: Very Encouraging Finish to the Year

In recent weeks, financial markets had begun to increasingly worry about inflation again. Whether it was natural gas prices rising again, the incoming Trump Administration's tariff plans, or fears around electricity prices exploding because of AI demand, markets had sent inflation expectations to their highest levels since April of '24.

Thankfully, December's CPI report did not play into that narrative. This year's report showed headline inflation increasing 0.4% (in-line with expectations for a 0.4% increase), and 2.9% year-over-year. This is a high rate on an annualized basis but predominantly due to energy and shelter prices (which we'll discuss more in a moment). This was in-line with expectations. Core inflation, which excludes food and energy prices, rose 0.2% on the month, or 0.23% unrounded. This was better than expectations of a 0.3% increase, with whispers indicating this could even surprise to the upside and round up to 0.4%. Year-over-year, core inflation ticked down a tenth to 3.2%, as the table below shows. This was a big sigh of relief.

Let's go through the report in more detail. As we usually do, let's look at the report through the lens of our three "buckets," as the Fed does: goods, services excluding shelter, and shelter.

First, let's look at goods prices. Goods prices for the month added +2 bps to core CPI. Notably, this included a +2 bps contribution from new car prices and +3 bps from used car prices. While this is down from the +7 bps these categories contributed last month, they're still elevated compared to their historical levels (zero, see charts below). Excluding car prices, goods prices contributed -3 bps to our monthly core figure of +23 bps. Used car prices bounce around and will likely continue to do so, but there is still not much reason to believe prices increase at problematic rates going forward, especially with rates where they are. All things considered, the goods category was pretty uneventful for the month (which is a good thing), and if anything, represented a slight headwind relative to where we expect to be, yet core inflation was still only 0.23% in December.

Used car prices decreased again in December, according to Manheim, and were up only 0.3% year-over-year, despite there being monthly volatility throughout the year.

Auto loans remain at significantly elevated levels, which has made auto payments significantly higher for the same price. It's actually pretty remarkable that we've seen the rebound in auto sales that we have (16.5M SAAR in December of '24 versus about 17-17.5M pre-COVID) considering this.

Core services, or "super core" contributed +8 bps this month, which is about in-line with the pre-COVID average of +7 bps. Two things are worth highlighting here: first, auto insurance contributed +2 bps of the month. This is below the worst of recent months (see second chart below), but still a tad elevated compared to the pre-COVID average (zero).

The second thing to call out, and this is only because it will likely be a source of upside pressure in coming months as well, is airline fares (see the third chart below). This month airfares contributed +4 bps, which is relatively in-line with recent months, but also well above COVID averages (this too was essentially zero). Given the issues the discount airlines are having, however, the supply / demand balance in the airline industry is tipping meaningfully in favor of the airlines (at least the ones that have survived anyways). For those who haven't booked a flight lately and need a sense of what I'm talking about, check out United or Delta's stock prices. Now look at their history on almost any time frame.

These companies rarely see this kind of appreciation, but it's definitely real at the moment, and consequently, it will likely be a source of upward pressure in coming months. Thus, while we shouldn't easily exclude this out from this month, this likely isn't a source of long-term pressure, as the airline industry has historically not been able to restrain itself from adding capacity during good times and driving fares back down as a result. The encouraging thing is actually that we were able to print a core services ex. shelter ("super core") number in-line with pre-COVID averages including an above average contribution from airfares (and car insurance, for that matter). There are undoubtedly always categories that contribute more or less each month, but that's kind of the point. We may be in a position now where category level analysis for super core becomes less relevant, and all the noise cancels out (hotels for example, were a negative contributor of 2 bps this month versus a historical average of, you guessed it, zero). Time will tell there.

Core services ex. shelter came in basically in-line with the pre-COVID average (+8 bps vs. +7 bps pre-COVID average) despite having airfares and car insurance still providing excess sources of upward pressure. This is encouraging and continues to demonstrate that our entire inflation "problem" remains with the CPI's calculation of shelter.

Car insurance continues to be a source of excess upward pressure on inflation, but significantly less so than in recent months. Things seem to have worked their way through the system here, for the most part, but we'll see.

Airline fares are volatile, but their contributions of late have been significant, especially relative to the pre-COVID averages. Considering the capacity situation in the airline industry, however, we should probably expect these kinds of contributions from airfares in coming months.

Now let's turn to shelter. Shelter contributed +14 bps of our core +23 bps in December, in-line with last month's +14 bps contribution as well. As the chart below shows, CPI's shelter contribution is coming down, but slowly, for reasons we've talked about in prior posts. Patience is the operative word here, but there's no reason to think this won't continue to come in. As I'll show below, if you make no other adjustments to core CPI except for shelter, and we use a "market-based" shelter index, core CPI inflation year-over-year was 1% in December.

If there's anything that remains "sticky" in the inflation world, it's the way the CPI calculates shelter inflation. It's not the data itself, but the methodology for how CPI chooses to cut that data that has made shelter inflation look sticky.

Most private market data is not out for December yet, but look at the gaps between what private market sources say shelter inflation is compared to what CPI's shelter index is telling you. CPI said shelter costs were up 4.6% year-over-year in December, but private market data sources will likely end up indicating that it will be a faction of that. ApartmentList's rent tracker, for example indicated rents are down 0.51% year-on-year in December.

Changing nothing else in core CPI but shelter, and using private, market-based data sources instead of CPI's shelter component, core CPI would have been 1% year-over-year in December (half the Fed's target). Using November's data, which is more complete, would have produced inflation below 1%.

Let's also look at inflationary breadth, which can help us assess inflation trends as well. If more categories are inflating at faster or slower rates, it can give us a better feel for inflation trends in the real economy, without the distorting effects of certain categories with higher weights in the aggregate indexes (like shelter).

As a reminder, the way we assess inflationary breadth is by looking at sub-categories in the CPI. I call these "Category Levels," simply because in Table 2 of the CPI reports each month, they're noted by numerical indents in the spreadsheet. Category 4 consists of 55 baskets of goods and services that make up 98% of the core CPI, while Category 5 level "indents" include 101 baskets of goods and services that comprise about 77% of the core CPI. Given Category 4 comprises 98% of core CPI, that deserves more consideration than Level 5, but both are instructive.

As the charts below show, category level inflation has been below aggregate inflation in the core CPI for most of the last several years, including the last several months. This is evidenced by the lines (which are average and median monthly price increases of Category 4 and Category 5 goods and services) being below the bars (which are the total core CPI increases each month) in the charts. For the month of December, for reference, Category 4 goods and services saw median and average price increases of 0.10% and -0.03% respectively, while Category 5 level goods and services saw 0.00% and -0.11% median and average price increases (all of this is on a seasonally adjusted basis).

We can then take this data and annualize it to get an even better picture of core inflationary "breadth." First I do one more thing, which is to average the last three months worth of data, and then annualize it. Below then is a chart showing 3 month average readings for aggregate core CPI (the bars), Category 4 goods and services (the blue line), and Category 5 goods and services (the orange line).

As you can see from the chart below, the lines have dropped significantly again to close to zero, after a spike up to 2% last fall. Two things are worth highlighting here: first, the lines have remained below the bars here for most of the last several years; second, given the lines are so far below 2% on the Y-axis, the category level data are telling you inflation in most of the economy is well below the Fed's 2% inflation target. Even the spike produced last fall still indicated inflation in the real economy was probably not much more than 2%, despite the aggregate index indicating otherwise.

Looking at median and average increases in larger baskets of goods and services can give us a good sense of inflationary breadth in the economy, and the good news here is that when we do so, it shows that inflation on median and on average has declined to levels even below the Fed's 2% inflation target. The above chart shows 3 month averages, annualized, of this category level data to make it easier to assess where we are relative to the Fed's 2% inflation target. Since the lines are again close to zero, it's telling us that inflation in most categories has significantly declined (to essentially zero, on a three month basis).

In summary, December's data showed yet again that our inflation "problem" is almost entirely shelter. Except it's not even the shelter data itself, but rather how CPI cuts and calculates inflation from that data that is telling us we still have an issue. Though it's been nauseating having to write about this month after month, this should continue to get better in coming months and years, as the CPI's methodology simply results in a lag. None of us probably wants to hear it, but as we mentioned earlier, patience may turn out to be the operative word in 2025.