Home > CPI, Tweet Summary > Summary of My Post-CPI Tweets (April 2019)

Summary of My Post-CPI Tweets (April 2019)


Below is a summary of my post-CPI tweets. You can (and should!) follow me @inflation_guy. Or, sign up for email updates to my occasional articles here. Investors with interests in this area be sure to stop by Enduring Investments or Enduring Intellectual Properties. Plus…buy my book about money and inflation. The title of the book is What’s Wrong with Money? The Biggest Bubble of All; order from Amazon here.

  • CPI in under 10 minutes. Gentlefolk, start your engines!
  • Start with the consensus: this month economists are calling core 0.18% m/m, and 2.1% y/y.
  • Remember that last month, core was 0.11%, in a downside surprise driven by pharmaceuticals and autos, while Rents were actually somewhat bubbly again.
  • However, Median inflation turned out to be 0.26%, pushing y/y to basically tie the post-crisis highs at 2.77%.
  • That should lead you to suspect that there were some ‘tail effects’ last month that could be reversed this month. So that would make me marginally bullish this number, all else equal.
  • Now, there’s talk about the fact that the BLS is changing its collection method for Apparel to use a direct feed from retailers rather than manual price-sampling.
  • Some people think the change in the method of collecting Apparel prices should depress Apparel, but I’m not really sure why that would be so unless there was some systematic bias in collections pushing prices higher.
  • If so, I’m not sure it’s showing. Apparel is -0.76% y/y. In any event, Apparel is only 3% of CPI so effect should be quite small. And apparel recently has been weak. So I’m not too worried about that. Famous last words, I guess.
  • Core prints at 0.1%, 2.0% y/y. But that’s not as weak as it looks. It was actually something like 0.148%, whereas market was looking for 0.18% or so…y/y is 2.04%. So both barely rounded down.
  • last 12 months’ core CPI chart. Just bumping sideways. We ought to be back to 2.1% y/y next month, as we drop off a weak April 2018 print.

  • Well, trust the bow-tied set, I suppose. Apparel -1.94% m/m, -2.2% y/y. I guess those manual price checkers were pushing prices up, after all. (?)
  • And CPI for Used Cars and trucks, second month in a row, weak at -0.38%. That’s lower than Black Book (which has been a much better fit than Manheim since last year’s methodology change) suggests it should be.

  • But more importantly and lastingly – rents remained firm, with primary rents +0.42% m/m and OER +0.32% m/m. That keeps OER stable y/y and raises Primary rents to 3.68% y/y.
  • Primary rents y/y. Not sure if this is an aberration because I don’t track market rents. Seems unusual for late in the cycle, but wage growth has been strong and supports this dynamic. But seems a bit strong.

  • Pharma bounced, rising to -0.39% y/y from -1.19% y/y. But the downtrend doesn’t seem terribly damaged.

  • Core ex-housing drooped a little bit, not surprisingly given the breakdown. Core ex-shelter is 1.10%, down from a 1.54% November high but still well above the 2017 lows of 0.53%.
  • Interestingly, like last month where Used Cars fell and New Cars gained, the same thing happened this month. Used cars & trucks went to 0.44% y/y vs 1.11%; New cars to 0.72% from 0.29%. A little odd, but just wiggles.
  • Although Medicinal Drugs re-accelerated slightly, Professional Services (doctors)) decelerated to 0.39% from 0.97% y/y, as did Hospital and related services (1.94% from 2.12%). As a result, Medical Care WOULD have decelerated but for Health insurance.
  • Always worth a reminder: health insurance in the CPI is a residual, since CPI measures only the portion of medical care that individuals pay directly. But it rose to 9.06% y/y from 7.66%.

  • This chart is why we like to ignore core and focus on median. Clearly a lot of left-tail stuff going on.

  • Primary rents y/y. Not sure if this is an aberration because I don’t track market rents. Seems unusual for late in the cycle, but wage growth has been strong and supports this dynamic. But seems a bit strong.
  • So, having said that…my early guess at median CPI is for +0.27%, which would push median to 2.85%, clearly the highest since the end of the crisis. We will have to wait a couple of hours for the official figure.
  • Four pieces. Not much change this month except in the last piece. Here’s Food & Energy.

  • Piece 2: Core goods. Dragged down by used cars, pharmaceuticals. Our models have this still going higher so I think these are one-off effects.

  • Core services less rent-of-shelter. Doctors, hospitals dragging this down. Be wary if Medicare-for-all proposals start to gain traction; if they do then I’d suspect doctors and hospitals would start to raise prices before their prices get fixed or cut.

  • Part 4 is Rent of Shelter. I’ve been saying forever that we’re not getting deflation because this isn’t about to fall off a cliff. On the contrary, it’s actually moved above our ensemble model.

  • So, here’s our ensemble model for OER. Primary rents are actually a [little] bit above our model. As you can see, we’re expecting a gradual slackening of rental pressures. BUT…

  • …but our model based on income (not shown) rather than home prices is actually calling for higher rents. You can argue that higher wages have helped produce these higher rents.

  • But if that’s the case, it means that when inflation is actually rising, looking back at home prices is NOT the right way to do it. Indeed, if the wage hypothesis is the driver then we’d expect to see a divergence in Primary and OER rents that leads shelter costs higher.
  • ..there’s no real sign yet that primary rents are accelerating way past OER, but it’s something to keep an eye on if rent inflation continues to surprise on the upside.

  • That’s all for now. Thanks for tuning in.

The upshot of today’s report is that while there are lots of small one-offs that are making sharp moves lower, and each of them has the potential to cause month-over-month mischief, the broad body of prices is remaining stable and/or edging higher. Regardless of what happens to apparel (Women’s and Girls’ Apparel, Men’s and Boys’ Apparel, and Footwear accounted for the three largest declines this month) or used or leased cars and trucks, housing costs appear to be moving higher.

It is early to be certain about this, but there have been anecdotes about faster rent growth in places and there are some signs (as in the rent chart above) that these pressures are diffusing more broadly. With strong wage growth, it wouldn’t be terribly surprising if there was more household formation (there has been, although not a huge amount) and more pressure on rents as people move up. How this resolves is key to the medium-term outlook. If higher wages help to push rents higher, and continue to put downward pressure on the number of existing homes available for sale (see below, source Bloomberg), then core inflation is simply not going to droop lower in the way the Fed expects it will. If, on the other hand, this is just a temporary rise, then the one-off declines in core inflation will eventually be joined by soggy rents and core will drift somewhat lower.

Either way, I see little chance that core or median inflation will even remotely approach the lows from the last cycle, even if we enter recession later this year as I expect. To get there, housing would need to implode again and the dynamics are simply too healthy for that at the moment.

Advertisements
Categories: CPI, Tweet Summary
  1. No comments yet.
  1. No trackbacks yet.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.

%d bloggers like this: