E-piphany

Here we are again, on the monthly CPI roller-coaster. Consensus coming into the day was +0.60% m/m headline, +0.33% core, pushing the y/y numbers to 3.7% and 2.7% respectively. The headline print will obviously be flattered by energy prices again, but the core number may seem surprising since we have been running between 0.196% (last month) and 0.313% (last July) for the last year. Why so high? To be honest, the core forecast seems lowish to me: this is the month where the shelter rebound is to take place. As a reminder, the BLS methods from the missed month of CPI in October implied no change for shelter, and that’s clearly wrong; because of the way the sample rotates, we had to wait 6 months to get the correction. That day is today. The consensus of good inflation shops seems to be that the correction will be worth 0.14% on Core, which makes you wonder…0.33% minus 0.14% would be 0.19%, so essentially economists are expecting new lows in m/m core? Seems curious.

Last preliminary point – the market over the last month has generally impounded higher inflation expectations (far left, inflation swaps). The sharp decline in 1y CPI is an artifact of the fact that over the last 30 days or so, a lot of good carry rolled off the front. In fact, that makes the 2y look all the more remarkable. Nominal govvie rates are higher (right column), but that’s a good bit due to higher real interest rates (third column) more than inflation expectations. Which is interesting.

Now for the actual data and the rain of charts.

The actual print for CPI was +0.64%, with m/m core CPI of +0.376%. So the economists were low, and even the swaps market was low as NSA printed +0.85% m/m.

The jump, as I said, is partially a repayment for the very low Oct/Nov numbers last year (there was no October, so this chart divides the very low 2-month change in November). Figure if the run rate was +0.25% or so per month, we were a cumulative 40bps too low over those two months. Not all of that was given back today; much of it showed up in December and January. But that is the context for the big spike on the right-hand side. It’s not as bad on the (estimated) median chart. My guess is +0.33% m/m for Median, but that’s almost certainly off by a little since the median category is going to be one of the regional OERs.

Core Goods inflation dropped to +1.1% y/y this month, while Core Services rose to +3.3%. In September, before the shutdown, the numbers were +1.5% and +3.5% respectively, meaning that we have made some progress – but not as much as we need to – on core goods and very little progress on core services. Bottom line is, the improvement it appeared we were seeing in CPI over the last six months was basically an artifact of the shutdown.

Here are the rent charts, with a ditto mark for the prior comment. The jumps this month aren’t really a hook higher; they’re correcting the erroneous y/y numbers that the last 5 months showed. We have still made progress – back in September these numbers were 3.39% and 3.76% respectively. Just less than we thought. The m/m increase in Primary Rents was +0.55%, bringing the y/y to 2.79% from 2.56%; for Owners’ Equivalent Rent the numbers are +0.53%, taking it to 3.3% from 3.1%.

Additionally, Lodging Away from Home rose 2.44% m/m. This category did not have the same problem that the Rents series did, or rather the correction happened previously – there is no 6-month rotating survey for Lodging Away from Home. Lots of shelter inflation this month!

Airfares were +2.82% m/m. That’s in services, but it’s significantly related to jet fuel of course. With this increase, airfares are now roughly in line with the increase in jet fuel. Of course, mergers of giant air carriers and the bankruptcy (Spirit Airlines) of regionals are both damaging to the competitive nature of airfares, so it remains to be seen how much that effects prices in the long run. It probably shifts higher  and steepens the chart below somewhat.

Here are the Four Pieces charts. Food and Energy +8.03% y/y. Core Commodities +1.13% y/y. Core Services less Rent of Shelter +3.28% y/y. And Shelter +3.26% y/y. These four pieces, in descending order of volatility, add up to the CPI and they’re each between 1/5th and 1/3th of CPI. The one we tend to focus on, rightly because it incorporates the feedback loop of wages to prices, is the “Core Services less Rent of Shelter” one, aka Supercore, and it is not looking as positive these days. Indeed, none of these charts are doing what they need to do if we want to see 2%!

Here is median wages – the Atlanta Fed Wage Growth Tracker – vs Supercore. Median wages have increased the last couple of months. You can’t really call it a trend change yet. But, this is the feedback loop. If you want 2%, you really need wage growth to be about 3%. No real sign of that yet.

Here is another way to look at ex-shelter inflation. Above we saw core services ex shelter; the chart below is core (goods+services, not just services) ex-shelter. That line is in dark blue. Shelter inflation is in light blue. If you’re really optimistic about shelter coming down to below 2%, then you can argue the rest of core CPI is only a bit above 2% (about 2.6%). But historically, as you can see from the chart, you wanted most of core to be below 2% while shelter ran a bit above it, if you were going to be in a placid inflation environment. So that doesn’t look quite right yet. Plus…there’s no sign shelter is about to drop below 2%.

When I say ‘no sign shelter is about to drop below 2%, the chart below is what I mean. It had appeared that rent inflation was dipping below my model. But it turns out, the model wasn’t high at all – the dip was the artifact of the shutdown. So rent inflation is 2.8%, the model is at 2.8%, and the model a year from now is at 2.8%. That doesn’t seem like the right sort of trend. I seriously doubt rents are going to do the heavy lifting getting inflation down to 2%.

By the way, let’s not sleep on food prices. They were up 0.5% overall this month. This chart shows the level in the top panel (note that food doesn’t mean revert like energy does), and the rate of change down below. The 2024 dip was a rebound from the 2022 spike, but we’re crawling back up. Note that unlike energy in the CPI, the food category isn’t just a pass-through of commodity costs. Higher packaging, trucking, marketing, etc costs are more important long-term drivers than the cost of the commodity – and in Food Away from Home, obviously wages matters too. Food is just another category – a major one, to be sure – that doesn’t look like it is placidly dropping to 2%. Although, since it’s not a core category, the Fed could theoretically ignore it. It’s harder to look through food price increases than energy price increases, since as I said food price increases don’t typically mean-revert.

Now for the good news portion of our broadcast. The combination of Trump RX and the new “most favored nation” policy on drugs is having some effect on the CPI for “Medicinal Drugs” (which includes both prescription and non-prescription drugs). Prices are in fact declining. I would argue it isn’t super dramatic yet, but drug prices are coming down.

Put this all together and the Enduring Investments Inflation Diffusion Index rose again – reaching 44, which is the highest level it has ever seen other than the 2020-2022 spike.

With this April CPI, we now know the USDi coin’s price through the end of June. The current USDi price is 1.0394. By the end of June, the price will be 1.05488. That’s 1.49% over the next 49 days, so about 11% annualized.

Furthermore, inflation swaps for May’s CPI are suggesting another hottish NSA CPI of +0.55%, which would annualize to 6.6% if it happens. After that, energy correction (we hope) should weigh on CPI but for now, it looks like a hot summer for USDi. You can mint the coin at https://usdicoin.com/coin.

Wrapping this up, the read is actually pretty easy. Inflation is not just in energy, but right now is fairly wide as the diffusion index shows. Some of that is related to energy…the price of diesel fuel affects trucking costs, which affects other goods prices…and some of it is related to the fact that wage growth is no longer slowing. Any way you look at it, as I said the read is pretty easy: the Fed obviously isn’t going to be tightening into an oil shock. But there is nothing here that gives them cover to ease into an oil shock either. Warsh inherits a pickle.