Inflation continues to be the central focus of the market. And within the broader inflation measures, rent has emerged as a key upward driver. So what's going on with rent inflation? Why is it still firming, even as some private sector rent measures show some easing? On this episode of the podcast, we speak with Omair Sharif, founder and president of Inflation Insights, who breaks it all down, and more. This transcript has been lightly edited for clarity.
Points of interest from the pod:
What's going on with used cars? — 4:48
Why official data can lag private data — 8:34
How big of a deal is rent in CPI? — 12:24
Private vs. public rent data — 16:33
The challenges of shelter methodology — 19:34
Time to watch healthcare inflation — 27:32
The challenges facing the BLS — 35:18
---
Joe Weisenthal: (00:09)
Hello and welcome to another episode of the Odd Lots podcast. I'm Joe Weisenthal.
Tracy Alloway: (00:14)
And I'm Tracy Alloway.
Joe: (00:16)
Tracy. Last month’s inflation numbers, I guess it was just a disappointment, right? Because two months ago it was this cool number both on headline and core and it was like maybe the turn is finally here. And then last month it was like, nope, straight back up.
Tracy: (00:30)
How many words did you write about why people should be focused on month-on-month versus year-on-year?
Joe: (00:36)
No, I still think that from an investor standpoint or just someone trying to understand the economy, clearly, like the sequential numbers are more telling. Tracy I resent the insinuation that you'd think I would write something...
Tracy: (00:49)
I would never insinuate anything. Okay. But you're right. So people were expecting inflation to start to slow down a little bit. And that's why there was lots of talk about why you should look at the sequential month-on-month data vs. the year-on-year. But what we saw instead was basically, anyway you sliced the data, it looked disappointing and ... were you gonna defend month on month again?
Joe: (01:12)
No. Of course not. I was just gonna say ironically the only measure that looked good was the completely unsliced headline data, because that was so dragged down. But the moment you did even the slightest bit of prodding under headline, it was like, ‘Oh my God, raging hot.’
Tracy: (01:27)
All right. Well I think the big takeaway from that number, other than it being disappointing, was the fact that we really see some of these price increases starting to spread from things like food and energy and more towards services. And services, as everyone is now finding out, is a big, big part of the core index.
Joe: (01:44)
And I guess everyone called this too, like I think last year they're like, ‘Oh, we're gonna have this big shift of services.’ And yeah, sure, goods prices will come down and bullwhip effects and inventories and all that. But now it's here and it's like, ‘oh man, this could be here for a while and it's not slowing down yet.’
Tracy: (01:59)
Yeah. And I think the big question is how long does it take to feed into the index and how long does it take to kind of go away? And there are different data points and there's been some discussion of this as well. There's private market data points, for instance, that show rents are starting to slow. So when does that feed into CPI?
Joe: (02:15)
Right. So various online companies, like a Zillow or something, they'll have a rent index. And this sort of gets to the question because first of all, the thing that [was] one of the big upward drivers of inflation in the last report was rent. Everyone, most people feel rent. It's a very salient category. There are some categories that maybe are more hidden. Rent is not one of them. Shelter is not one of them. But then yes, there is this thing. So it's like, okay, we have these private measures that seem to be rolling over a little bit, but the numbers in the CPI keep going up. So is this a case where the CPI is just lagged? Like is it bad data or is it people are misunderstanding the relationship between the official government numbers and what some of these private surveys are saying?
Tracy: (02:54)
Well today we are gonna be digging deep into those numbers.
Joe: (02:59)
Absolutely. So let's get right to it. We have the ultimate guest for digging into inflation numbers and he knows more about like what these numbers actually mean and how they're derived than anyone. I think you've talked to him several times in your reporting on like the minutia — like when you did like mayonnaise reporting, it's like, okay, where is this in which index? And how much of this is like soy oils versus condiments?
Tracy: (03:24)
The way we measure inflation never ceases to amaze me. And there's just so much to say about the actual construction of the indices and things that people don't normally talk about, but we probably should.
Joe: (03:34)
Let's talk about them. We are gonna be bringing back to the show a past guest. Omair Sharif is the president and founder of Inflation Insights and he will answer all our questions about why some numbers are going up. Maybe some, hopefully they go down. So Omair, thank you so much for coming back on the show.
Omair Sharif: (03:49)
Thanks for having back
Joe: (03:49)
What's going on? When are the numbers gonna start turning down? I thought inflation was transitory. I was like on team transitory, now I look ridiculous.
Omair: (03:59)
Well I think probably, you know, the turn of this year is what I'm thinking is we're gonna start to see the monthly rate, especially in the core, start to really kind of come off. I mean we've been kind of stuck around 0.5, 0.6 every month on the core, pretty much for close to a year. We haven't gotten a lot of relief, but I think that relief is coming from a few different areas, hopefully in the next couple of months.
Joe: (04:21)
It doesn't matter how long it takes, I'm going to claim victory.
Tracy: (04:25)
In five years if inflation comes down? Great. Okay, well maybe we can talk about one of the things that people expected to start coming down, and it hasn't, at least according to last month's data and that's used car prices. And this was one of the big drivers of inflation actually, you know, going up over the past year or so. Why haven't prices come down? This is what everyone was expecting to happen.
Omair: (04:48)
Yeah. So used car prices, like a lot of things, you know, when wholesale prices are going up, they adjust very quickly on the way up. When they are coming down, they take a little while to come down. And so what we've seen honestly all year long is that wholesale prices are down very, very sharply. You wanna look at Manheim, Black Book data and JD Power, which is source data actually for the BLS. All of these things that are down, you know, 10%, 11%, 12%, over the last six, seven months.
Retail on the other hand has been kind of roughly flat because a lot of sellers are not really pulling down their prices. But we are starting to see that change. And really over the last six to seven weeks, I would say that that is starting to adjust.
So model years, you know, two-year models, three-year models, those prices, retail prices are coming down. You can see this in the Black Book retail index for example. Prices were down about two and half percent in August. So all we're really waiting to see is that data translate into the CPI index of used cars. And I think, you know, September is a good point where that might enter, but I think October is probably the month you really wanna focus on, September, there's a lot of adjustments that are happening in September. But September/October I think is when you really are gonna start to see the retail stuff on the used cars index start to come off in the CPI. And that's gonna be a big boom in terms of, you know, getting that core index lower.
Tracy: (06:08)
How do you actually go about, because this is your day to day business, how do you go about trying to measure that lag between, you know, what we see in the market and what we actually see showing up in the CPI data?
Omair: (06:21)
First of all, you're looking at these market indices, you know, whether it's Manheim or Black Book. And you're essentially them mapping them at the wholesale level against the retail price index for the CPI. Now obviously wholesale, you know, typically you'll find its strongest correlations will be with wholesale changes feeding roughly about two months later into the CPI. Now what's really interesting is like, before the pandemic, this was pretty much a constant, roughly a two-month lag. So eight weeks later, whatever happened wholesale typically showed up in the CPI.
What the pandemic did was completely throw these lags off to the point where, you know, prices started moving at the wholesale level and obviously they're going up, and that immediately fed into to retail. Like, there was no lag whatsoever. And if it went up 5%, 6% of wholesale one month, guess what? The CPI was going up, 5%, 6% of used cars the next month.
What we're seeing now, I think is a little bit of a reversion to the old lag of about two, three months. And that's why I think, you know, we've seen wholesale come off the last several months and people are saying, ‘Well, wait a second, the CPI used cars indexes, you know, [it’s] kind of unchanged or maybe up a little bit.’ So I think it's just the issue’s that the lags are kind of back to what they were. This is sort of like gasoline. You know, when gasoline prices go up at the wholesale level you feel it right away and you see it right away at the pump. When they go down, it's a very gradual decline, right? You're kind of pocketing that extra margin. And I think dealers are doing the same thing and have been for a while, but I think that jig is up, and should be over the next couple of months.
Joe: (07:52)
So let's talk a little bit more about this and, you know, I'm starting to like feel a little bit better about always being wrong on everything. Because I just look at these measures, right? I look at trucking. I look at cars, I look gasoline. All these lines are going down. They should show up. Can you talk a little bit about why these lags exists? What is the difference between some of these measured prices and the government prices and do you have any theories or sort of explanations why past lags got disrupted so that the gap between private survey measures of prices versus what showed up in government data did not have the same temporal relationship as it used to pre-pandemic?
Tracy: (08:31)
Whoa, temporal is a good word.
Joe: (08:33)
Thank you!
Omair: (08:34)
Let's go back to used cars. In terms of the pandemic, when it hit, what you had was basically a complete disruption in the used cars market. So it wasn't just that people demanded used cars because let's say they were moving to the suburbs or whatever. The biggest shift was you had a natural seller in the used cars market — rental car companies — who became a net buyer of used vehicles. And we really haven't seen that before. And what they did was they scooped up all the zero- to three-year models that they could, to sort of replenish their stock. You know, things came back a lot stronger obviously in 2021 that people thought.
So take a company like Avis, they had about 450,000 vehicles in their fleet in 2018. That got down to about 290,000 by the summer of 2021, right when demand was moving. So they couldn't replenish them for manufacturers. They ended up going to the used cars market. So what you found was price increases in the zero- to three-year bucket, which is all they’ll buy, those price increases skyrocketed. If you're a dealer trying to buy a car, you can't find these vehicles, right? So now you need to go out buy four- or five-, six-year olds. So all across that age curve prices spiked.
And I think the issue is just the magnitude was so large that you couldn't really wait to price them out on retail, right? So you were paying way, way more than you were used to paying. You had to pass that on quickly. And so I think we saw those lags disrupted for that in particular. Now on the way down, wholesale's getting a lot cheaper. You can still sell it for a bit more on the retail side. So again, just like gasoline, you kind of pocket the margins as long as you can on the way down. And I think that's what you're seeing.
The other item I would actually mention where honestly I'm pretty stunned as to what's going on, is really furniture prices. You know, you talk about trucking rates going down import prices of furniture going down, inventories have jumped. All the big box retailers are telling you, you know, Walmart, Target, we've got too much of this stuff that we're discounting heavily and yet the CPI is going up about 1% every single month. And there, typically, you would start to see the stuff come through pretty quickly.
I've honestly been waiting since March or April for furniture to really slide and it just continues to sort of defy expectations. So there I don't have as good of an answer in terms of what's happening, but we know all the signs are pointing to the fact that they should be dropping. And the industry will tell you that. I mean, you read any furniture industry trade publication like I started doing when I was wrong on furniture for a while you know, Furniture Today magazine will tell you they're preparing for recession. They've got too much stuff, they're discounting. Orders are declining, and yet in the CPI it’s not showing.
I will say one other thing to think about is some of these samples in the CPI are not that large for furniture. If you're talking about bedroom furniture, they may only have a couple of hundred quotes in the entire sample versus, you know, thousands for something like rent. So the smaller sample, the larger, you know, the chance that you'll have some, some errors you might miss, some of the price changes in a particular month. So that could also impact the lags that occur in the CPI.
Tracy: (11:49)
So I mean, it is true on the whole that goods inflation has been going down. So I think it fell from like 10.7% last year to 7.1% last August. And meanwhile, services, as we mentioned, is starting to pick up. And services, correct me if I'm wrong, but I think it's something like triple the weight in the core index, something like that. Can you walk us through A) why do services get that much weighting and B) how significant is it for the core index that it's now going up?
Omair: (12:23)
Yeah, so I mean the bulk of that weighting is shelter, it's rent and it's owner's equivalent rent. You know, in the core CPI, 40% of the entire core is just rent and OER combined. And so within services, you know, the bulk of the weight is coming from shelter. So that's really what's driving that overweighting, if you will, to services relative to goods. The fact that it is going up so dramatically, you know, that's obviously been an issue for the core. I would really say probably since early spring when shelter really started to accelerate. Now, you know, one thing I want tomention, is that I think the shelter story honestly is something that most people knew was gonna happen coming into this year, right? You mentioned some of these private market indices, like apartment lists and Zillow and so on, they were showing these huge gains in rents, you know, late last year, last spring, last summer. So we knew that this was likely going to enter in the CPI this year.
The question was always about magnitude. So whether it was gonna be up 6% this year or 7%, which looks like we're headed for that 7% number. So to me like that's not really a surprise on the services side. I think most people who track this stuff closely realize, ‘hey, rents are gonna be up a lot this year.’ Probably somewhere in the 6%, 7% range. It's the other part of services or the non-shelter service stuff that I think is the more interesting part of the story. And there what you'll find is a lot of people talking about how wages are driving those services up. You know, how all of these other costs in the non-shelter services, those are the sticky elements of inflation.
And until that stuff starts to roll down, you know, it's gonna be really hard to get core inflation down. I would actually sort of counter that a little bit by arguing that a lot of what you've been seeing, and this has been true since really last probably I would say fall, is we've had a lot of oscillation in that non-shelter services component. And that's mostly because of sort of the economy reopening and closing in kind of fits and starts. So summer of ‘21, if you remember, airfare started to jump very significantly. People started traveling a bit again. So a lot of what was driving services at that point was actually things like airfares and hotels. It wasn't medical care services, it wasn't recreation services, you know, it was really personal care services. It was these sort of reopening categories if you will. Then you had, you know, I think Delta was later that year, prices for those categories fell.
The non-shelter services inflation actually decelerated very sharply. So basically what you've seen up until really pretty recently is just this quarter to quarter oscillation that's been going on in the non-shelter services index, really just reflecting kind of the economy reopening, and then slowing down. We got that same dynamic this spring by the way, when airfares spiked and now they've been down the last few months. And so the non-shelter stuff is kind of moderating a little bit.
Joe: (15:41)
So let's talk a little bit more on rent specifically for two reasons. A) because it is such a big part of core CPI, but also it's one of the most salient. People feel it and people complain, certainly in New York, everyone is aware of just like how brutal the rent market is, but also elsewhere, you mentioned that we sort of had reason to think that this number was coming in part because the private surveys were flagging this several, like, maybe even a year ago, the Zillows and all this. Sot hat raises one big question which is like, is the data stale? And at a time when people are worried about is the Fed gonna create a recession? Is it operating on old data that's not as timely as what the private sector surveys are showing? Is the public data stale?
Omair: (16:33)
So I wouldn't necessarily say it's stale. I would just say that it measures something different than what these private sector indices are measuring. So most of these private sector indices are measuring new leases. So when you think about moving into a new apartment, you sign a brand new lease. That's what they are capturing. They're counting that change in the rent for that unit with a new tenant versus whatever it rented for the last tenant. And that's true of all of, most of these Apartment List and Zillow and so on. And so it's really just one segment of the market that those private measures are capturing.
The CPI is capturing the entire rental market. So it's not just people who are looking for new apartments who are signing new leases, it's also people who are renewing their lease. And it's also people who were currently renting and still on the same lease they were, you know, five, six, seven, eight months ago. So they wanna capture the entire market versus just a slice of the market. So in that sense, I don't think it's stale.
Now that said, you know, when you think about is the Fed operating on old data, we do know that it lags, right? For this very reason, the BLS doesn't capture turning points in the market as well as these private sector measures, right? If something is changing in the marketplace, the way those new leases are changing is gonna be a much better indicator of what's happening than the CPI will.
That's true, but it's not as if the Fed number one doesn't watch the other private sector measures. It's not that, you know, they don't understand the lags. I mean, if I understand the lags in rent and other people do, I promise you the folks at the Fed do as well. So I don't think that they are, you know, sitting here working on these sort of lagging indicators, if you will. Because they're capturing a huge amount of data to look at what's happening in the shelter. And they also kind of see where shelter’s likely headed, right? These private sector measures have started to roll over the last, you know, depending which one you wanna look at, four to seven months they've been slowing down quite a bit.
Joe: (18:36)
So another thing related to housing and the cost of shelter, which again seems important because of the weight and just how much, how important it is for the public. We have seen a clear downtrend in anything related to home buying and home purchasing. Is there a relationship or a stable relationship between activity and the home purchase market? The price of a house, the price of a monthly mortgage, which has shot up, if you're buying a house today versus a year ago, and then what feeds through into rent prices?
Tracy: (19:11)
Can I just say that was gonna be my next question?
Joe: (19:13)
Tracy and I, we've been working together so long, we always do this.
Tracy: (19:16)
We keep asking the same questions, but can I just tag onto that? So one thing I've heard is there are some people who say that like interest rates going up could end up increasing the pressures on rent because more people decide they're not going to buy houses, they're gonna stay where they are or keep renting an apartment, things like that.
Omair: (19:34)
So on that latter point, yes, that's, that's very possible. You know, if it's getting too expensive to get a mortgage or you can't find a house to buy, you renew your lease or you, you know, are moving into new apartment. That certainly can actually push rents up in the short term until supply does eventually catch up. But yeah, that's very possible. We've seen that happen before. In terms of the idea of, you know, the housing market and home purchase activity, it really is kind of what you're talking about, which is the knock-on effect on the rental market. That's really the way it's gonna feed through into the rent index because, you know, contrary to popular belief, house prices don't play any role whatsoever in the CPI at all, even though [there’s the] owner's equivalent rent index, you know, it's not intended to measure house prices.
It basically is using the contract rent data that they capture and sort of, you know, rejiggering it a little bit to come up with with OER. But no house price goes right into the index whatsoever. The mortgage interest stuff, you know, it used to actually be in the CPI prior to 1983. Because they just, it was a very different methodology back then. And so when rates were moving higher and the cost of servicing your mortgage moved up, that price actually was reflected in the CPI back in the day. But in 1983 there was a lot of different problems with it. And they ended up switching out to this new method of rental equivalents in 1983. So now that doesn't really play play much of a role, again, other than the knock-on effect on the rental market even was happening in house.
Tracy: (21:06)
Is that how Volcker defeated inflation? By removing mortgage rates from CPI?
Joe: (21:10)
It's like, ‘oh, we wanna raise rates to fight inflation. But our current measure of inflation includes mortgages. We better change the rules because otherwise our rate regime won't help us at all. It's a little weird.
Omair: (21:21)
So actually in that instance, because of the rate increases mortgage interest costs in the CPI skyrocketed. And actually they're pushing inflation higher. So even though he was boosting rates at the same time you would think, okay, higher rates you should lower inflation. In fact, inflation was moving higher partly because mortgage interest costs were so much higher. And there was a lot of problems with the idea of mortgage interest costs. I mean, they knew about mortgage costs and the problems with sort of putting it into a cost of living index.
A lot of issues that people have with CPI, whether it's rents or or other indexes, is really about the concept of how you design it, how you think about what you should be capturing. And that fundamentally gets back to the idea of, you know, what is the purpose of the CPI? And it's intended to be a cost of living index.
And you know, they knew back, I think it was in seventies, they had papers at the BLS saying, ‘look, we need to get away from this mortgage interest costs because it doesn’t really fit the way that we're supposed to construct a CPI. You know, you could do a whole separate episode on that, but I think the short version is in 1983, they decided to say, ‘Hey, we've been talking about this rental equivalents method for many years now, and we think it's the right way to do it.’
And by the way, I will just say, very recently National Academy of Sciences basically put out a report that said, you know, here are our recommendations for improving the CPI in the coming years. And they talked about looking at, you know, these private measures of rent as potentially trying to incorporate them into the CPI. But they said up until then, the best measure that we have is really the rental equivalents method that we we use today.
Tracy: (23:00)
Let me ask a slightly less provocative question other than how we measure or don't measure mortgage interest in CPI. So historically, one of the reasons we focus on rents is because people feel them. They're a big component of the core index, but also because rents and wages tend to be tightly linked. And I think there's concern that as rent inflation accelerates, are we gonna see that knock-on effect into wages? What are you seeing there?
Omair: (23:31)
As you said, it is a pretty tight fit. I mean, basically I would say labor income and rent growth are pretty tightly correlated. You know, and again, I think as rents have gone up, they've correlated well with this improvement in wage growth. One of the interesting sort of tidbits is that even though rents are rising, a lot of people who are reupping their leases are actually, or signing new leases in, you know, sort of more professionally managed apartment buildings, so more of your large multi-family unit buildings, are actually showing that their incomes have increased pretty significantly over the last two years.
So even though we talk about, you know, the idea of a lot of people getting priced out because rents are rising so sharply, people who are signing new leases and having to provide the paperwork from their bank statements or their employment information are showing that incomes have actually also increased pretty significantly.
And so I think, you know, as you start to see wage growth decelerate a little bit, which is already starting to happen at the margins, you know, people who manage these apartments sort of, they get this kind of real time flow of what labor income looks like. And I, I think that's partly also why you ought to see rents start to decelerate is because they're not raising rent 6% when these labor incomes are only growing, let's say it's 3%,
Joe: (24:51)
Where's that data from? That the cohort that is signing new leases is actually seeing wage gains that are keeping up with rent?
Omair: (25:00)
Yeah. So that comes from RealAge. And that's another, you know, large sort of private market provider of everything from rent data to all sorts of information on multi-family buildings. And so they've been tracking this and sort of publishing, you know, stories on this for the last, I think about 18 months or so. Just this idea that even though rents are moving higher, people are able to afford those rents because their incomes are rising alongside those rent deals as well.
Joe: (25:28)
So one other question on rent before we go off it, and again, it sort of connects to broader housing questions. A lot of people I think in the last couple years when rates were low bought houses as investment properties and maybe don't wanna sell right now, in part because there aren't a lot of buyers who are excited about the sticker shock of what a monthly mortgage now costs them. Could this bring more rental supply to market in your view? It’s like, ‘Well, I can't sell it, so I'll rent it.’ And could that have a dampening effect on rents?
Omair: (25:58)
Yeah, very possibly. I mean most indices don't track single-family rentals. The only one that I'm aware of that does is it's either Zillow or CoreLogic. It's one of those two has a single family rent index where they do track what rents are specifically for that type of of rental. And yeah, obviously if you can't sell it and you bought as an investment property, it makes quite a lot of sense given that, you know, demand is still pretty robust for rentals. Vacancy rates have only barely started to edge higher from the lows that we saw, you know, even six months ago. So there's still quite a lot of demand out there. So it makes sense to do that at this stage. And yeah, hopefully that can help bring at least that one segment of rent of the rental market down. And by the way, that is also captured in the CPI when you're talking about rent, it's not just apartments. They also do capture single-family rentals in that as well.
Tracy: (27:12)
So we've been talking a lot about rents which have been pushing services up higher, but are, to the point you've been making, are expected to start to decrease sometimes soon, or at least the rate of acceleration will start slowing down. Talk to us about another big change that you see potentially on the horizon, that has to do with healthcare.
Omair: (27:32)
Yeah, so this is another one that I think is coming very soon. It's ggonna help everyone looking for that transitory inflation story to kind pop up again. So as I mentioned, the biggest part of the core CPI is 40%. The second biggest component is medical care. That's worth just about 11% of the core. And for the last year, medical care has been rising at about roughly 0.5% each month. Which means it's been adding about five basis points to the core change every single one. That's been very steady, it's kind of like clockwork, pretty much all of last year. Starting in October, that index is gonna turn negative and it's gonna turn negative in most months over the course of the next year.
And so what was a pretty constant source of a boost to the core every single month is actually gonna turn into a relatively decent drag on the core. And, you know, it's not because medical care is getting cheaper or so on. This is actually just one of these quirks in the methodology that you kind of have to be aware of. And it comes very specifically from the health insurance index within the broader medical care gauge.
Joe: (28:52)
Why is that? What's the change that's coming? What is going to switch from pushing up to being a drag?
Omair: (28:59)
Yeah, so the story is basically that, well, first of all, health insurance is updated once a year, typically in October. It used to be September. Last year, it was October. This year will be in the October report again. But this data lags by almost about a year. So the BLS takes its data, and this is from an official source, the National Association of Insurance Commissioners. So they put out a big report on how much in premiums is being collected, how much is being paid out, and how much is retained by the insurer. So this is kind of the holy grail of this data set. Unfortunately, it doesn't come out until about 10 months into the year. And so what we're really capturing this October is gonna be activity that happened in 2021.
And so what's gonna happen here is that if you think about 2020 and during the pandemic, people put off things like elective surgeries. They didn't go to the doctor because people didn't want to be in waiting rooms with other people who might have Covid, right? So there just wasn't a lot of utilization of healthcare services in 2020. Premium income continued to increase, but the benefits paid out actually declined. The way the CPI captures health insurance is by looking at the change in these retained earnings for insurers from one year to the next. And you know, very quickly the reason they do this is because it's really hard to price health insurance from one month to the next because policies are changing all the time, right? What a policy will cover will be could be changing quarterly, monthly, you know, each year. The risk factors that go into the policy can change.
So when the BLS is pricing any good or service, they want apples to apples from one month to the next. And if something changes, they want it quality adjusted. But in healthcare and in health insurance, they found that's just way, way too hard to do. So the roundabout way, this indirect way of capturing the price to you, the consumer, of health insurance, is basically what does it cost the business to offer you health insurance? If the cost of administering health services or health insurance is rising, you'll probably see that in your premiums. And so the way they capture this is by looking at these retained earnings and how they're changing from one year to the next, So during the pandemic premiums kept rising, however benefits paid out to people went down quite substantially because of Covid and the lack of utilization of healthcare.
So what you saw was a huge spike in retained earnings. And what that meant for the CPI was that in October of 2021, which reflected this 2020 data, health insurance jumped by 2% in the month of October, which means that basically since you only update it once a year, it's effectively gonna print right around 2% every single month like clockwork until the next update.
Fast forward to 2021. People started going back and taking care of these elective surgeries and utilizing healthcare much, much more than they did in 2020. Premiums didn't really change too dramatically. So now you have way more utilization of healthcare than you had in 2020, in 2021. And so then those retained earnings dropped on a year-over-year basis. So what's gonna happen now when you update it is that you're going to have a very large drag.
So health insurance, which has been 2% a month pretty much for the last year, is very likely to print around -4%. You know, on the surface you say, well, 2% to -4%, that doesn't really sound like a lot, but if you kind of put it into context, it actually is kind of dramatic because number one, in the month of October itself, that alone is worth almost a seven basis point swing on just the core CPI. So if you're forecasting, let's say the core, maybe, you know, 0.4% in October, you're probably looking at something that's more like a 0.3% instead just because of this move in health insurance. And you know, that's kind of the difference between a 5% annualized rate and, you know, something that's more like a three and half percent rate.
So that's a pretty big gap. The other issue here, on a year-over-year basis, health insurance currently is about 25% year-over-year because of the steady, you know, 2% march every month. When the next report comes out in September, which is sort of the last before it, you know, turns negative, we'll probably hit about 28% year-over-year. Once this -4% comes in in October and it stays there for the bulk of the year, by September of 2023, health insurance I suspect will go from +28% to about -40% year-over-year. That swing is worth almost about 80 basis points on the core CPI. So you have an index that's worth just over 1% of the entire core CPI, that by itself will subtract almost a full percentage point from core inflation over the next year. That’s pretty dramatic
Joe: (33:54)
Tracy, this is so crazy to me because the whole point of like measuring this stuff and then monetary policy is like countercyclical, and this huge component as you just described, it has no, there's no actual like economically-cyclical impulse part of it.
Tracy: (34:10)
Well, this was actually gonna be my next question and Omair that was absolutely fascinating. And the thing about qualitative adjustments was something that I only really discovered this year. So I didn't know, the BLS, you know, if they're looking at the cost of a refrigerator, for instance, will take into account technological advances on the cost of the refrigerator if it now comes with, I don't know, WiFi connectivity or something — a blockchain-enabled fridge. And they will factor that into the CPI.
But I mean, this gets to one of the major criticisms of the indices themselves. You can kind of see what they're trying to do. So it's difficult to measure qualitative improvements in things like healthcare insurance. But on the other hand, it does lead people to look at these things and go ‘Well, what are we actually measuring here?’ And isn't it weird that the cost of living as measured by the CPI, which includes rent, healthcare, food, energy, whatever, can change just because of the way this one thing is measured, retained earnings versus the way we measure goods and things like that? What do you say to that criticism?
Omair: (35:18)
It's fair to make those sorts of criticisms, I guess I would say a couple things. One is that, you know, this is never a static process in terms of the methodology. The BLS is always trying to improve upon whatever it is that they're doing. A good example of this is something like new vehicles. Just in April of this year, in fact, they introduced a brand new methodology for capturing the price of new vehicles. So before they used to go to dealers, figure out what was selling, you know, try to capture those prices. Now they're using a massive data set from JD Power, which captures, you know, essentially real live transactions that are occurring. So they've updated that quite significantly to really reflect kind of the conditions on the ground for people who are buying new cars. So there's always this sort of, you know, goal to improve upon the methodology. So that's number one.
Number two is, you know, you kind of do the best you can with what you're given. And by that I mean that a lot of these things are subject to things like budget constraints. You know, when we talk about rent, for example, the BLS, if you survey a unit, let's say in January and you say, ‘Hey, how much are you paying in rent?’ You come back to that unit, you don't come back to it in February or March or April. You come back to it in July — six months later. Part of the rationale is because, you know, rents don't change a lot in terms of the contract. So six months seem like an adequate amount of time, but the two other reasons are, one, respondent burden, right? If I'm knocking on your door every single month asking you what your rent is, you might be less willing to participate in the survey.
But the other is also, there's a budgetary constraint in terms of, you know, sending people out into the field to capture a lot of these data sets. So all of these things sort of constrain what the BLS can ultimately produce. In this particular instance for health insurance, you know I can sort of understand a bit more of the criticism, but the issue here is really the data is just lagged 10 months. We can't do anything about that. I mean, the data that they are getting is from the National Association of Insurance Commissioners. And if you think about capturing all of the premium, all of the claims that are paid by all of these health insurers, you know, it just takes, for the last year, it just takes a while to put those numbers together.
So this is just something where the BLS just has to wait on the data that they're capturing. Again that data set is, you know, effectively like the bible for health insurance data, right? And so in this instance, they can't do anything. They just have to wait until that's produced. So I think, you know, in some instances I get the criticism. I just think you, you've got to understand that they're working within a number of different constraints. And you know, sort of take that into account when you're thinking about criticizing them for particular approaches in terms of, you know, how they construct an index
Joe: (38:13)
Omair ,can we do a live event with you one day where people throw out a CPI category and then you like, on the fly, explain it. No, I’m serious. People would love that. Explain how it's constructed. Cause we could talk forever on every category. Seriously, can we do that someday?
Omair: (38:29)
Yeah. There's, you know, there's 243 individual components in the CPI, and I think I've got most of them down. So we could probably do that.
Tracy: (38:38)
‘Other fats and oils including peanut butter.’
Omair: (38:42)
I didn't say we can do the PPI but…
Tracy: (38:45)
Oh wow, you actually knew that was a PPI code versus CPI. That's very impressive.
Omair: (38:51)
I think we talked about that. One final thing I just wanna make sure on this health insurance stuff, is that part of the reason why I think it is pretty important is because this data set, the official report is coming out, I think in about a week. I tend to use a separate source, which, you know, captures quite well. But basically, you know, my feeling is that folks who are either trading inflation or who just generally follow inflation are pretty unaware of this change that's coming. So this is gonna be something that, you know, at the margin is going to help the Fed month-over-month for the next year, along with, I think, the coming decline of used cars. So really Q4 potentially is shaping up to see some lower core inflation for not just used cars.
Tracy: (39:34)
That makes Joe so happy.
Joe: (39:35)
I'm gonna spike the football at the end of Q4. Omair Sharif, thank you so much for coming on. Fascinating conversation. Always love chatting with you. Always learn something and we'll have to have you back again soon. Thank you.
Tracy: (39:47)
Thanks Omair, that was great.
Joe: (40:01)
I love talking to Omair. I always learn so much. Setting aside the fact that he gave me, you know, he like throws these little like bits of red meat for team transitory, I just actually like learning about this, like I had no idea how they captured health insurance. That's so interesting to me.
Tracy: (40:14)
So I think I have maybe three major takeaways from that. One, it's just crazy how much of the market and our daily lives are linked to the construction of this one index and, well, I mean it's multiple indexes, but PPI, CPI and how it works. Think about all the payments, Treasury-linked securities, things like that that are linked to CPI. And so much of it depends on the individual construction.
And then the second takeaway, you know, what he was saying about the time lags and how Covid kind of messed those up, I think that’s a really good way of looking at why there's been so much confusion over inflation. And then thirdly, this is something that I'd heard before, but the resource constraints on the BLS in terms of assembling some of this data and trying to adjust it, I think that is maybe an underappreciated factor over the past couple years.
Joe: (41:06)
Well, and especially like some of the stickier prices within goods that, like, of course these should come down, right? Because we have every big box retailer saying, we have tons of inventory, the housing markets slowed down. So it's like all kinds of reasons to think, yeah, we should be seeing some deflation in furniture. It's not happening. But then he says maybe they only could track a couple hundred in the survey.
Tracy: (41:27)
Well, right. If your survey respondents are like the big box stores that have pricing power or have pricing power for a while, then it'll be sticky.
Joe: (41:36)
There’s so many. We should at some point do an episode on when they changed the rules of inflation, because it's crazy to think that like, you know, 40 years ago if they raised rates, that mathematically raised measured inflation because interest in mortgages was included in the CPI, which is also like kind of maybe intuitive to a lot of people.
Tracy: (41:55)
Well, it's one of those things like you can see why they would do it, but on the other hand, it also seems odd if you think that CPI is supposed to measure the cost of living.
Joe: (42:05)
Fascinating stuff.
Tracy: (42:06)
Yeah, we could talk about this for a long time. Shall we leave it there?
Joe: (42:08)
Let's leave it there.
You can follow Omair Sharif on Twitter at @fcastofthemonth.