Archive for April, 2021

Once Again, You Ain’t Getting No Coke

April 27, 2021 5 comments

For a long time, I’ve held the opinion that the notion of “anchored inflation expectations” was an absurdity. For one thing, we have no good way to measure inflation expectations: market-based measures don’t reflect consumer expectations, and survey measures are nonsense that mostly reflect an availability bias (i.e., changes in small, frequently-purchased items, especially gasoline, have a much larger impact than large, infrequently-purchased items). There are lots of other biases in inflation perception, some of which I enumerated and discussed in a scholarly-ish article almost a decade ago.

It isn’t that I think that people don’t have inflation expectations, or that they are ‘wrong’ in some sense. It is just that the notion that they are “anchored” is something that is completely unmeasurable and so hypothetical. But many economists believe that hypothesis is necessary to help explain the break in inflation models around 1992-1993. I think there are better explanations for that break, which don’t require assuming a can opener.

However, recently I have started to reconsider whether there is a way in which behaviors concerning inflation are at least sticky. This is not to say that I think this necessarily has a role in inflation modeling (importantly, because there’s no good way to measure it), but I have definitely seen anecdotally some behaviors that can only be explained by figuring that consumers and producers become at least conditioned to expect low and non-volatile inflation. (Note, if I’m right about it being a conditioned response rather than an anchoring with respect to “strong central bank messaging,” it is useless in explaining the 1992-93 inflation model break because my hypothesis is that it takes a long time to happen).

My thoughts derive from some direct observations I have of actual producer/supplier behavior, from customers of mine and their suppliers, over the last couple of years but especially in response to the latest spike in raw materials prices. When I first began this sort of risk and pricing consulting a few years ago, I was struck at the attitude that one of my customers seemed to have – the customer behaved as if it was a commodity producer facing extremely elastic demand curves, such that they were very convinced that if they raised prices at all, they would lose a huge amount of their business to suppliers in China and India. Their customers of course reinforced this notion by responding to questions about price by saying that “lower prices would be good.”[1] But their product was both higher quality and shorter lead time than that of the competition; yet, they priced it as if only price mattered to their customers. The important point, though, is that they were conditioned to believe that any increase in price would destroy their business.

Their attitude wasn’t unique. Until recently you could see that behavior all over. To a surprising degree, diners with printed menus (with pictures!) have tended in recent years to have prices pretty much hard-coded and printed onto the menu. Your local barber probably changed prices every five years, at most. And every change was usually accompanied by an explanation to the customer about the need for higher prices for one reason or another. It is a strange dynamic, very different from what you see in an inflationary regime, such as in this classic scene from the movie Caddyshack (released in 1980):

Tony D’Annunzio: Give me a Coke.

Danny Noonan: One Coke.

[gives Tony a bottle of Coke and 50 cents]

Tony D’Annunzio: Hey wait a minute. That’s only 50 cents.

Danny Noonan: Yeah, well, Lou raised the price of Coke. He’s been losing at the track.

Tony D’Annunzio: Well I ain’t paying no 50 cents for no Coke.

Danny Noonan: Oh, then you ain’t getting no Coke!

Instead, in this era of low and stable inflation it looks and feels like suppliers have learned “anchored” behaviors. “I can’t raise prices; all of my customers will leave.”

Fast forward to the COVID crisis.

Input costs for many producers have skyrocketed. For example, the price of polypropylene has roughly tripled since the lows last summer. Yes, at least part of that is temporary. Up until about six months ago, when movements in input costs induced changes in profitability for producers they would hold the line on prices, lest they lose a bunch of business, and watch their margins decline. Ultimately, when input costs retraced, they would enjoy wider margins again. Price changes were artificially muted because the supply chain dampened price fluctuations. The producer absorbed those costs because of the perceived elasticity of end product demand.

However, the volatility of input prices recently has been such that producers couldn’t absorb all of the costs into margin and still remain viable. Some of the suppliers and competitors to the client I mentioned above did indeed hold prices as long as they could, before eventually passing them on with great apologies. I advised my client, though, to pass cost increases through immediately. Some in management wanted to label the increase a “surcharge,” but that again is an apologetic way to adjust prices – and one you do need to justify. Instead, they jacked up their prices to maintain their margins, and braced for the worst.

And lo and behold…nothing happened. Some customers wanted an ‘explanation’ for why the price of the Coke went up, and some customers complained that the Coke was cheaper elsewhere. Ultimately, my client did just as much business after the price increase as they did before the price increase.

What they learned, and what lots of suppliers, restauranteurs, and others are learning, is that demand is not as elastic as they had thought; that everyone wants a lower price but they’ll pay for the value of the product if they need the product. That, with more money in the system, raising prices a little doesn’t hurt business very much at all. If you need a Coke, you’ll pay the fifty cents for a Coke even if you wish it was still forty cents (and part of the reason you will pay without much complaint is because you just got $1,600 from the government for no reason at all).

Watching this behavior is what makes me wonder about the anchoring of inflation expectations. Again, I think of this as a conditioned behavioral response. By the same token, though, people can unlearn these conditioned responses. We are all conditioned now to put on a mask when going into a restaurant, but we will (hopefully) unlearn that behavior and the new conditioning will be to not put on a mask when entering a restaurant. That’s what’s happening now, I think, to suppliers in many industries. I think they’re all surprised, but I also think they’ll remember. It’s just one of many reasons I think that regardless of the path inflation takes over the next decade, it will likely be both higher on average and more volatile as well. Take away the “anchor” and the ship tosses about more, and moves with the tides. I don’t really know of a good way to model this, and I am also fairly confident that the Fed will not recognize that the anchor has been slipped (which means bad things in expectations-augmented Phillips Curve models!) for a while. To be clear, I don’t think the lack of policy response to the un-anchoring matters much because I don’t think the Fed had very much to do with the conditioning in the first place and expectations won’t become “re-anchored” merely because of the central bank’s messaging. I guess I don’t really think the central bank’s messaging means as much to consumers as they think it does! In any event, it takes time for conditioning to take hold, but less time I think for the conditioning to be broken.

[1] One of my first recommendations was that they stop asking this question. What do you expect customers to say, that they want higher prices? The question is whether they will not buy your product at a higher price, and asking their opinion on that is a really bad way to find out the answer.

Categories: Uncategorized

Summary of My Post-CPI Tweets (April 2021)

April 13, 2021 2 comments

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, issuers and risk managers with interests in this area be sure to stop by Enduring Investments!

  • Good Morning #CPI observers! Prepare for what is potentially the most entertaining #inflation figure in a while.
  • Before I get started, let me first note that I’ll be a guest on ( with @OJRenick at around 10:20ET this morning. Tune in!
  • Today’s walk-up is a little different. I usually try and focus mostly on the y/y numbers because the m/m numbers are an accumulation of random distributions around 280 other numbers. That is a lot of noise compared to signal and so I don’t like to forecast monthlies.
  • However, on a y/y basis the noise tends to cancel so it’s a clearer reading. Median CPI is even better because it lessens the impact of the tails.
  • This month, however, and for the next few months the y/y number is a distraction. We KNOW it’s going to jump a lot because the comparisons to March, April, and May 2020 are super easy. So instead, we want to focus on what happens to the monthlies.
  • I warned about this back in February in “The Risk of Confusing Inflation Frames.” And now…here we are.
  • So looking back at the last couple of months, we see that the core CPI figures were soft. Last month, core CPI (but not median CPI!) was soft because of surprising movements in goods, outside of housing. It had been goods pressing core inflation higher so that was surprising.
  • Turns out that some of that was (probably) due to the fact that the weather prevented the BLS from surveying certain prices. So we’d expect a little catch-up from last month’s +0.10% core, just as a null hypothesis.
  • Some of the places we are pretty sure to see strength are in autos, apparel, and the travel categories. Used car prices are nuts. But in the bigger picture, there are a lot of shortages out there and they all push prices the same way.
  • I talked about some of those shortages in my article at the end of March. How Many ‘Shortage’ Anecdotes Equal Data?
  • There are shortages in autos (due to semiconductors as well as lower fleet sales into the used car channel), packaging, cotton, containers, rental cars, Uber drivers, other goods…and shelter.
  • In shelter, rents have been artificially soft because of the eviction moratorium, which has made realized rents decelerate while asking rents are rising rapidly with home prices. That divergence is unusual and it’s due to the eviction moratorium.
  • The Biden Administration just extended that moratorium (was due to expire end of March) so that catch-up will come later. However there are SOME signs that rents are improving anyway. I’ll be looking for that. Rents were not as soft last month as they had been recently.
  • The economist consensus is for a core CPI m/m of about 0.2%. That seems low to me with all of the potential upside disturbances, and has got to mean that economists are expecting further shelter weakness. I don’t.
  • The market doesn’t either. Interbank trading of the (headline) price number implies about 0.1% higher than the economists expect. Most of that in core presumably. I would not be surprised in the slightest at +0.3% core.
  • We will see. Remember, the Fed doesn’t really care – and they’re working hard to tell you that you shouldn’t either. Eventually, the market will win. But not for a while. It will be late 2021 before the dust clears on the base effects.
  • So keep an eye on those underlying pressures and don’t get distracted by the y/y fog of war. I will talk today in terms of y/y figures, out of habit, but rest assured I’m watching the small ball too.
  • Thanks for coming along today on this crazy ride. Good luck! 6 minutes to print.
  • OK, core came in at 0.34% m/m, so quite a bit higher than estimates. y/y rose to 1.646%…so ALMOST rounded to a 2-tenth miss on the y/y figure.
  • Note in that chart, they’re not y/y. There’s no base effects there. In fairness, we probably should combine the last two figures, and get something like 0.22% per month, but that’s still faster than the Fed would like. Except they don’t care.
  • So Core Goods jumped back up to 1.70% y/y, where it had been 2 months ago before dropping to 1.3% y/y last month. Collection issues. Core Services up to 1.6%.
  • Primary rents +0.15%; OER +0.23%. Not as soft as a couple of months ago, but not overly strong either. Lodging Away from Home was +3.84% m/m, which pushed the Housing category to a +0.34% m/m rise…same as core, weirdly.
  • Apparel fell again. That’s a bit odd. Apparel had been doing well partly because cotton imports from part of China were being held up at the ports…maybe that’s lessening now. Anyway Apparel isn’t a big piece.
  • Pharmaceuticals: +0.08%. Doctors’ Services: +0.28%. Hospital Services +0.63%. First time I can remember them all three being positive in a while! Softness in Pharma is still surprising to me.
  • Doctors’ Services highest in years (y/y).
  • Hospital Services, despite this month’s jump…not so much.
  • Back to used cars. Part of what is happening here is that rental fleets shrunk last year so they are providing fewer cars to the used car markets. Part is the semiconductor shortage making new cars expensive. But Black Book says…this has a lot further to go in months ahead.
  • Ah. Core CPI ex Shelter jumped up to 1.61% y/y. Yeah, I know I said y/y. But that was at 1.7% last February BEFORE the COVID slide. Arguably it means price pressures are higher now than before COVID, and CPI is being held down by rents.
  • This isn’t from the CPI report but a reminder of what is happening in rents. If a landlord is unsure of being able to collect the rent, it goes in a zero. Doesn’t take many zeroes to lower measured rent. And the number of zeroes is higher when the gov’t says you can’t evict.
  • Other COVID categories: Airfares +0.44% m/m (fell 5% last month!), Lodging away from home I already mentioned +3.8% (-2.3% last month). Motor Vehicle Insurance +0.85% m/m.
  • New Cars, interestingly, was flat. That’s odd – there’s clearly a shortage of semiconductors so maybe this is more a situation of you can’t get ’em so the price doesn’t change? I’d expect that to rise going forward.
  • Car and truck RENTAL: +13.4% (SA) m/m. Here’s the m/m and y/y, which is now up to +31%. If you can’t buy ’em, you can try to rent ’em. Remember how I said fleets are smaller?
  • Now, Median CPI giveth and Median CPI taketh away. Hard to tell because median category will probably be a regional OER, but m/m will be probably 0.2-0.22%. Median y/y won’t change much b/c base effects were mainly from a few small categories with large moves.
  • That warrants further comment: the fact that we didn’t see a GENERAL deceleration in prices, but a very focused one, should make you wonder about output gap models. Most of the economy wasn’t in deflation. Hotels and airfares were though!
  • Only two core categories with more than a 10% annualized decline this month: Women & Girls’ Apparel (-28%), and Infants’ and Toddlers’ Apparel (-22%).
  • On the gainer side, tho: Car/Truck Rental as noted, Jewelry/Watches (+80.7% ann’lz), Lodging AFH (57%), Motor Vehicle Insurance (+47%), Men’s/Boys Apparel (+35%…hey!!), Misc Personal Svcs (+16%), Motor Vehicle Maintenance & Repair (+12%).
  • Core goods & Core services. Both rose, and remain atop one another. How long can goods stay elevated? Port traffic is improving, slowly. But materials prices remain stubbornly high and global trade remains fractious.
  • ok, gotta wrap it up and get to makeup for my appearance on @TDANetwork at 10:20. KIDDING, no makeup. You can dress a monkey in silk but it’s still a monkey. Anyway, I’ll do the four-pieces and then conclude. Will put out the diffusion indices later.
  • Piece 1: Food & Energy. No surprises here: it was expected to jump as gasoline prices continue to recover.
  • Piece 2: Core Goods. Back to the highs.
  • Core services less Rent of Shelter. This still remains bizarre to me. But medical finally showed some life this month and there’s sign of pressures in the PPI there so maybe it’s coming. Hard to see an uptrend here though unless you turn it upside-down.
  • Finally, Rent of Shelter. It seems it may be done going down, and there’s a lot of catch-up to do when the moratorium ends. But the last 2 months of rents have been more normal.
  • So at this hour, 10-year breakevens are +1bp and stocks are flat. Because the Fed doesn’t care, and the punch bowl remains. I guess that’s about the summary here. The base effects are going to obfuscate whatever is really happening underneath.
  • BUT, what is happening underneath (per the chart of core-ex-shelter) appears to be price pressures that are certainly no smaller than pre-COVID. Are they temporary? How will we know? If the Fed says they are, and are wrong…bad.
  • If the Fed says the pressures are NOT transitory, and are wrong, and over-tighten, that’s also bad – but for employment. And here’s the thing, this Fed has said repeatedly that full Employment is their main goal. So errors are designed into the system to be inflation-enhancing.

Here’s the summary of the main points today. Ex-housing core inflation is back at the level it was prior to COVID. Housing is artificially depressed because of the way the BLS accounts for rents (which is reasonable, since someone who isn’t paying has certainly decreased his cost of living), and asking rents tell a totally different story. But since measured rents are soft, it means that core isn’t low right now because of COVID categories: it’s low right now because of one thing, really, and that’s rents. If realized rents converge upward to asking rents, you can tack another 0.7%, 0.8%, 0.9% or so onto core CPI.

Inflation is already higher than it “should” be coming off the greatest global economic contraction since the Black Death. And that’s without consumers being truly unleashed. But the Fed has adopted an asymmetric policy stance, because they very publicly feel that the risk of higher inflation is something they ‘have the tools to manage’, whereas they believe they have some sort of moral obligation to make sure everyone is employed. I don’t want to draw too many parallels to prior hyperinflations because that’s not what I’m looking for, but the current asymmetric stance is very odd for any policymaker who learned history and knows that one of the reasons that Weimar Germany printed so many marks was because they believed having everyone employed and paid was absolutely crucial, and so they ran massive deficits and printed money to pay for them.

This is why the Bundesbank has always been willing, ever since, to rein in inflation even if it meant short-term pain in labor markets. They remember that the best route to maximum employment in the long run is to maintain a stable pricing environment. As recently as the 1990s, the Fed (Greenspan at the time) would regularly say that. It is no longer the core belief of the FRB.

The Fed believes they have the tools to rein in inflation, the knowledge about how to calibrate them, and the will to use them, but at least for the next 6 months they will wave their hands vaguely at ‘base effects.’ After that, if inflation is higher than they would like once the base effects are past, they’ll vaguely wave their hands and say ‘average inflation targeting.’ It it going to be a very long time before central bankers willingly hike rates without the market forcing them to do it. And before that, there may very well be a showdown where the Fed decides to defend the longer-term yield environment and implements Yield Curve Control. These actions and possible actions have very different implications for stocks and bonds depending on the path, especially with equities pricing in a goldilocks environment. Get ready for a bumpy year.

%d bloggers like this: