Odd Lots - Omair Sharif Explains How Inflation Measures Really Work

Episode Date: September 2, 2021

When people think about what inflation is, they might first think about some broad index like the CPI. What does the the CPI really tell us? And how is it constructed? And how much does its rise and fall relate to the state of the macro-economy? On this episode, we speak with Omair Sharif, a longtime veteran of both the buy and sell-side, who is now the founder and president of Inflation Insights. Omair knows inflation index construction better than anyone, and he walks us through what's happening right now, and how he thinks about forecasting where inflation will go.See omnystudio.com/listener for privacy information.

The news doesn't stop on the weekends. Context changes constantly. And now Bloomberg is the place to stay on top of it all. Hi, I'm David Gurra. Join us every Saturday and Sunday for the new Bloomberg this weekend. I'm Christina Rafini. We'll bring you the latest headlines, in-depth analysis, and big interviews. All the stories that hit home on your days off.

And I'm Lisa Mateo. Watch and listen to Bloomberg this weekend for thoughtful, enlightening conversations about business, lifestyle, people, and culture. On Saturday mornings, we put the past week's events into context, examining what happened in the markets and the world. That on Sundays, we speak with journalists, columnists, and key political figures to prepare you for the week ahead. Join us as soon as you wake up and bring us with you wherever your weekend plans take you. Watch us on Bloomberg Television.

Listen on Bloomberg Radio, stream the show live on the Bloomberg business app, or listen to the podcast. That's Bloomberg this weekend. Saturdays and Sundays starting at 7 a.m. Eastern. Make us part of your weekend routine on Bloomberg Television, radio. and wherever you get your podcasts. Hello and welcome to another episode of the Oddlots podcast. I'm Joe Wisenthal.

And I'm Tracy Allaway. Tracy, do you remember, I think it was probably like last June or last July, like over a year ago now? I think we were talking to Vitor Constancio. And I think you asked like one of the most important questions like at the time and maybe still like that we've had on this podcast. I struggle to remember things that happened a month ago, let alone 12 months ago. Everything is sort of blurring into this one long stretch. Yeah, that's fair. What did I ask?

Well, you asked him whether economists or anyone these days had a cogent theory of inflation. Oh, yes. Do you remember? And at the time, like, you know, there wasn't much inflation going on. I think we were probably still in something resembling deflation or disinflation, not long after the initial shock. But now, you know, fast forward to August 2021 and official inflation measurements are pretty elevated relative to recent history. And I think that question, and I think his basic answer is no, although I don't want to quote him on that. That question of like whether economists understand inflation has probably never been, is extremely top of mind these days.

Totally. And I mean, I think I would agree with you on the no part. But of course, there's a really big irony right now, which is. Right when the Fed changed to the flexible average inflation targeting regime after, you know, more than a decade of undershooting the 2% inflation target, as soon as they did that, now we seem to have possibly transitory inflation, but like certainly more than they would have expected a year ago. And, and yeah, like, I think it just illustrates that no one seems to have a very good handle on what exactly causes price increases. or decreases. Right. So even prior to the new framework, there's always been this idea that, okay, we'll use, and this came up in our recent episode with Neil Cashcarry of the Minneapolis Fed, that, okay, we'll use inflation as our gauge, as our speed limit to know when we have, say, hit

maximum employment or full employment. That's true under the old Fed. That's true under the current Fed. Now we have elevated inflation, but it's like, oh, this doesn't really count because it's transitory and it's related to shipping and we know that doesn't have anything to do with employment, but there's always probably going to be a story to tell, which sort of calls to mind whether these are sort of like useful things. Of course, no one predicted this sort of like, are very few people accurately predicted the timing and the degree of elevation of the current inflation.

So I think we're still sort of like back to square one. We don't know how long this will last. I don't even think there's a wide agreement of like what really would be transitory of what would be worrisome of what would spur the Fed to act sooner than maybe markets affect. And so I think that question that you asked, Constancio, several months ago, remains really the sort of the key thing right now. Yeah, I agree. All right. So today we're going to be talking about inflation and really what it is and how it's measured and whether it's even possible to forecast it accurately or how we should be thinking about it right now. And we have literally the perfect guest. We're going to be speaking with Omer Sharif. He's currently the founder and president of a new shop that he set up called

appropriately enough for this episode, Inflation Insights. But he has a long track record. He's prior to that a buy side strategist at the asset management firm Millennium. He's been an economist at various shops, including Sokgen, RBS, and so forth. So we're going to talk about whether there is a a cogent theory of inflation and how to think about it right now. So Omer, thank you so much for joining us. Thank you having me. Well, what's your answer to Tracy's question? Do a, do economists have a useful or cogent theory of inflation that works in practice? I think the short answer to that is no. You know, all you have to do is sort of look at the fact that there's a massive academic literature that's basically just devoted to forecasting inflation.

and coming up with various types of models to figure out what the inflation process is. And there's an equally large literature talking about why we're so bad at forecasting inflation. So I don't know that there's a cogent theory. It seems to sort of change based on where we are kind of in the cycle. And there's different ways of approaching, you know, how you want to forecast inflation. There's sort of the top-down modeling approach that a lot of academics use. and I think sort of what's more kind of in favor now, something that I started doing, you know, well over 10 years ago,

which was kind of more of the bottom up approach to forecasting inflation. But ultimately, I don't know that there's a coaching theory that really can explain the inflation process over, you know, let's say the last several decades. You sort of try to understand it based on where you are, I think, in the cycle. So what is it that makes inflation so difficult to grasp? the idea of, that you just pointed out, that it basically depends on where you are in the cycle. And so your framework or the dynamics that are underpinning prices are kind of changing constantly.

Yeah, I think that's exactly what it is. It's just, it's a constantly evolving process. And, you know, one way to think about it is that if we're trying to forecast it, and let's say we just want to think about where it's going to go over the course of the next year, you know, there are all sorts of approaches you can take in terms of. terms of modeling. But some of the simplest approaches actually work the best, which is simply to say that, you know, if I want to forecast inflation over the course of the next four quarters, I might just use the average of the last four quarters. And most often than not, that will actually perform better than trying to come up with some models that, you know, Phillips curve type models,

for example. Because inflation persistence is actually a pretty big key, I think, in trying to understand the dynamics within inflation. And that persistence is, is, is, it varies across time. And that is one of the keys really in trying to think about it is, you know, if I were to tell you that the core CPI has been between one and a half to two and a half percent for the last 20 years, those types of models that are what we call naive models work quite well, because if inflation has been around that for the last 20 years, pretty good odds that, you know, for the next year it'll be somewhere in that range. But that persistence varies in the short run. it varies, you know, even within a decade.

And so trying to capture that kind of time-varying nature of inflation persistence is really what everyone's striving to do. And that's why certain models perform, you know, really well in certain decades. And they completely collapse in the next decade. Well, so this gets to something that we talked about in the beginning, that whether we're talking about the Fed's new framework. And we should point out, we're recording this August 23rd. By the time people listen to this, it'll have been the one-year anniversary.

of Jackson Hole, where they laid out, and so we make it some new speeches on this, but where they laid out the new framework last year. So whether it's the new framework, the flexible average inflation targeting framework or the old framework, I'm not even sure what that was, both are premised on this sort of like Phillips Curve thinking, that there is some inherent tradeoff and that there is some inherent speed limit or maximum employment and we'll know we'll get there, not by some number of employment level, but by inflation readings. And if the nature of inflation sort of changes all the time, maybe decade by decade or some other time interval, is that going to be a folly for the Fed to think that inflation or like, is there any reason to think that Phillips Curve thinking or a Phillips Curve framework will be a useful guidepost for the Fed? It's hard to say. I think the thing is that the feds, you know, their time horizons really, if you think about it, is essentially about three years, right? We get about three years of forecast from the Fed.

Fed within the SEP. And so, you know, if you're thinking about inflation changing over 10 years or 15 years, that's less of an issue for the Fed. So can the Phillips curve work accurately for them within a three to five year horizon? Sure. And, you know, there are times that it has actually performed relatively well. So, for example, kind of the late 70s, when you ran regression, sort of these, you know, using output unemployment gaps, Phillips curves actually turned out to be kind of useful. But, you know, the Fed will always tell you that they have a suite of that they look at. They will look at everything from core inflation to, you know, trim means and median CPIs. So I don't know that they're as reliant on the Phillips

Curb as they used to be. And I think they've kind of spelled that out for us over the course of, you know, the last year. It's not clear that the Phillips curve really works. And I think Powell's kind of ditched that approach. And it's, you know, it's kind of back to the old adage of saying, you know, we'll know it when we see it and we'll kind of wait for the whites of the eyes of inflation before we decide to move on policy. So you mentioned all the different inflation stats or figures that the Fed can look at. And of course, there are, I mean, there's probably dozens of different measures of inflation that are based on hundreds, if not thousands, of different baskets.

I guess my question is, how much does your interpretation or does one's interpretation or thinking around inflation actually depend on the measure that they're looking at. And how do people go about choosing which measure is most relevant in a particular time? Well, I think, you know, there are basically two main measures is the way that I really look at it. You know, there's obviously the CPI, and then there's the Fed's preferred measure, the core PCE. Are there other measures that you can look at, absolutely? But typically, they tend to be kind of variations on those two. So, you know, the trim mean and the median CPI and the median PC are just variations and different ways to kind of approach those same baskets.

And, you know, maybe you're taking some of the volatility out from the top and the bottom. But you're really sticking with those two main baskets. Which one you want to really focus on. You know, I think it depends. If you're the Fed, they've already made that decision for us. You know, it kind of makes it easy. We're going to be focusing on the core PC if you're thinking about monitoring. policy. But obviously for the markets, what matters is the CPI and what matters to the CPI because

that's what it goes into pricing tips. So it sort of depends on who you are. You know, if you're the Fed, obviously, you're going to be concerned. And obviously the two are related about, you know, roughly something like 70, 75 percent of the core PC is actually just passed through from the core CPI. So those to me are really the, at least in the U.S. are the two main metrics that you want to focus on. Well, let me ask you, you know, if economists, you know, don't have a great track record of forecasting inflation, they don't have great models for it. So we have these things that, okay, if I talk to Tracy, I say, let's talk about inflation. And we want to know where inflation is. We'll look up some indices on the terminal.

We might look up CPI, core CPI, core PCE, et cetera. Are the concepts of inflation indices themselves cogent? In other words, we're aggregating all these different prices and trying to arrive at one number. And right now, I think the number, you know, CPI, it's a little bit over 5%, maybe 5.4%. I don't remember exactly. That's down from a recent high. But is there some true information contained in that headline number? Or is it just a number that gets spit out when you go through the work of adding up all of the thousands of prices that go into it?

Yeah, look, I mean, at the end of the day, all of this stuff is a construct, right? And there's certain things in there which I think do a great job of reflecting reality. And if we think about energy, that's a very simple one. You know, everybody knows what they're paying at the pump for gasoline. And within the CPI, you know, you're capturing these movements and energy prices. Those are pretty straightforward. The same thing is true of when you think about, you know, car prices. These are items that are pretty easy to capture and, you know, they do a pretty good job sort of representing reality. Where you start to get, we're starting to lose people, frankly, is when you kind of get into some of the sort of the price index theory and sort of the number theory where you start to talk about imputations and, you know, how to handle missing prices. And, you know, owner's equivalent, rent is a great example. There's been a debate about this for, you know, pretty much the last 40, 50 years about how should we really capture house prices? What is their role in an index? And the CPI is very clear. We don't want to, we don't want anything to do with house prices.

Why? Because we consider it to be an asset. We're trying to measure what you pay out of pocket as a consumer. And so these are the kinds of debates that I think, you know, when you're paying, when you're seeing, you know, home prices go up 20, 30 percent, you care less about what statusations are arguing about when it comes to, you know, imputation and so on, you just know that if you want to buy house, you've got to pay 20% more than maybe you did last year. And so, Pete, you tend to kind to lose the public when you're getting into those sorts of weeds in the inflation data. But I think for the most part, these indexes do a pretty good job of capturing what's going on in terms of the pricing environment around us. I'm June Grasso, inviting you to join me for the Bloomberg

Law podcast. Every weekday, we help you make sense of the legal stories that shape the nation and the world. Listen for complete analysis of the biggest court cases, the latest actions from Congress and regulators, and the legal moves driving the markets. From corporate law to constitutional law and from state courts to the Supreme Court. At Bloomberg Law, we go beyond the day's headlines. We speak with top attorneys, judges, scholars, and policy experts to break down what the rulings really mean. We do this every weekday, then bring you the best conversation. in our daily podcast. Search for Bloomberg Law on YouTube, Apple, Spotify, or anywhere else you listen. On the East Coast, listen as you start your day. And on the West Coast, catch up in the evening.

That's the Bloomberg Law podcast with me, June Grosso. Subscribe today wherever you get your podcast. Can we spend a little bit of time on owners' equivalent rent or OER? Because as you mentioned, this is a source of big controversy. Anyone looking at the housing market, you know, over the past few decades, will say that house prices have gone up, rents have gone up, certainly relative to one's income. So why is it that it seems difficult for the inflation indices to capture that? Is it just, is it just that they've made a conscious decision not to include it or to include it in this very specific way or is it something else?

So we actually used to include it. So prior to 1983, the CPI, for example, they took what was called the asset approach. And so that included, you know, the price of a house, it included everything related to your mortgage interest costs, your property taxes and so on. And even though from about the late 50s to about 1983, they did it in this manner, they knew from really the early 60s that this was not the right approach. And this was just a conceptual issue. the CPI. All this had to do was to say, what we're really doing here is we're capturing an investment piece of the house. And we really want to capture just the consumption aspect of it. And so they knew for, you know, 20 plus years that even though they were doing in this manner, this was not what they wanted to actually be doing in terms of the CPI. And finally, around the late 70s, they started doing more detailed work on how to get around, you know, eliminating the investment piece of it and focusing just. on sort of the shelter part of it. And they made that change in 1983. And, you know, another part of

the reason for making that change was, you know, as you recall, interest rates were kind of all over the map in the late 70s and the early 80s. And that introduced a tremendous amount of volatility in the CPI to the point where it sort of became, I don't want to say useless, but so volatile that it was really hard to get any kind of sense of what was going on with underlying inflation. and you had everything tied to it, you know, cost of living adjustments. Wage negotiations were tied to it. And one year it was up, you know, 10, 11 percent. The next year might be back down to 3%.

So it was sort of losing its significance. And so this decision was made to work on trying to implement this owner's equivalent rent index, which finally came into play in 1983. So that is a conceptual issue. It's just underlies sort of what the CPI is really meant to do. The second part of this, I think, is that, you know, when you look at, for example, Zillow, CoreLogic, and you're seeing rents up 70%, where you're seeing Case Schiller saying house prices are up 20%. One of the reasons you don't see it in the CPI indexes to that magnitude is because there's just very little turnover in the CPI sample. So every month, only about 10 to 15% of the sample actually represents new renters.

And so what you see in list prices, you know, that might be moving very quickly when market conditions are changing. That shows up over the course of about 12 to 18 months in the CPI because you just don't have that same kind of turnover that you see in these indexes that look at just, for example, listed rents. So part of its conceptual issue, but the other part is just the methodology and the way that the sample works, you're just never going to have the magnitude of the changes that you see in all these sort of, you know, private. private sector indexes. So like if we're talking about something like buying apples or buying milk or buying gasoline, everybody buys those things every day or every week all the time. With rent, not only do very few people actually sign a new lease every month, but also many of the new leases that people actually sign are with their current landlord.

And so they probably don't get the full market rent because those don't adjust as fast. And so, you know, looking at this, okay, this is one of the big questions right now. We know that headline inflation has come down a little bit. Used cars seem to have stopped going up. That's a big factor. But everyone's like, okay, OER is coming. OER is coming. We see the market rents on places like Zillow and so forth.

Those are shooting up. Why don't you give us what, you know, sort of you just explained the theory. What are you actually seeing in practice when you look at the data and how much is rent and other attempts that the CPI or that these indices used to capture? shelter, how much upward pressure are they going to put on the measures in the months and years ahead? Yeah, so we, you know, we have seen both of these measures, both right now we are actually bottom out over the course of the last several months. And you're starting to see price increases in the major metro areas. And so even, you know, places like New York, Los Angeles, Chicago, San Francisco, which are still down pretty sharply year over year. It looks like on a monthly basis,

they've finally kind of begun to stabilize a little bit. And, you know, it's important. important to kind of understand when you talk about the CPI that a lot of these rent indices, what really matters is where these so-called class A cities, which are the big ones with populations over two and a half million, what they're doing, because the weights on these cities is incredibly large. So if you looked at just three of them, New York, Los Angeles, Chicago, that is 20% of the entire rent index from just those three metro areas. So where they go matters quite a lot for the overall index. But we are starting to see these places stabilize and move up. But I think there's something to keep in mind here about, you know, this whole story about

shelters coming and how we are just going to go up and so on. Number one, people again are looking at these private market rent data and they're seeing 7%, 8% growth. We've never seen anything like that in CPI. I'm hard for us to think, you know, we'll see anything like that over, let's say, that of the course of the next year. Will rent go up? Yes. But, you know, don't forget before the pandemic, we were running around 3.5%. We're about 2.5% right now in rent and OER combined. So, you know, even if you move up a full percentage point over the course of the next year, you'll be kind of right back where you started in early 2020. But let's assume for a moment that, you know, we move up to 4.5%. So another 2 percentage points from where we are today. What that

basically means is you're looking at overall core inflation rising by roughly about another 80 basis points. You know, rent's got about a 40% weight. You go up two percentage points. That's about 80 bits on on the core CPI. Now, that sounds like a lot. But you mentioned used cars earlier. They're adding over 130 basis points to the core CPI right now on a year-over-year basis. That's almost certainly going to come off. So even if a Wii R goes up and rent goes up, you know, a couple of percentage points over the course of, let's say, the next year, 18 months, that's almost certainly going to be offset to a great extent by, a lot of these things that we're seeing now that we, you know, continue to think are transitory. That's going to upset a lot of the upward pressure you're going to get from shelter, I think,

over the course of the next 18 months. So it's kind of important to keep that in perspective because, again, 4.5% is really where we've peaked in the past. And even if we get a bit higher than that, you know, year over year, 40% on use cars is not going to stick. You know, that's going to potentially more than offset what we see out of shelter. So since we're talking about the current environment in terms of inflation, maybe it's worth asking, you know, when you hear the term transitory inflation, what does it actually mean to you? Because we've had at least one Fed official come on and talk about how they sort of regret using the term transitory inflation. And other people have said, you know, maybe it would have been better if the Fed was talking about narrow inflation versus more broader inflation. or manageable inflation versus unmanageable inflation. So what does that term actually mean to you, transitory inflation? So to me it's just about, you know, how long that rate of change continues to sort of accelerate.

So, you know, use cars going up year over a year from zero, basically, which is where they were pre-pendemic, to now of 40%. the question is how long can we sort of not just how long will that persist, but can it continually go up and sort of the rates we've seen over the last six months. That's kind of the way I think about it. That's, I think, the way that Chair Powell has sort of explained inflation as well is it's this process of can we see continually this rate of growth sort of accelerate year after year after year. And that's what we're sort of looking at for these components, but obviously much more sort of in the short term. But I do think there's a couple of ways to think about this transitory question. And if you're trying to figure out, is the inflation I'm seeing now transitory or is it

going to continue or will it spread out to other components, there's, I think, a few ways to kind of think about that. One, for example, is just to simply look at the dispersion within the CPI. So, you know, what share of components are seeing price increases today versus price decreases? And also, what does that look like on a weighted average basis versus history? So you could have a lot of components, for example, rising. But if combined the weight of those components is not that high, it really may not matter for kind of the longer term inflation picture. Another thing that I like to look at is momentum within the core CPI. So here what I want to look at is, you know, what is the share of components that are either accelerating or decelerating within the core CPI?

And here you can sort of, you know, normally you would look at, for example, the 12-month change in the year-over-year rate of specific components and see if that's picking up steam or losing steam. And that, once you wait those sort of changes, gives you a sense of kind of the underlying momentum that's really sort of driving the aggregate core number. And then one final thing, which I think is pretty important right now, especially since we are sort of comparing everything to kind of the pre-pandemic time, is you've got to kind of keep a close eye on where. the price level is today for certain components versus not just where it was, let's say, in February 2020, but where would you expect it to be today, you know, given the pre-pandemic trend? So are we overshooting that or are we undershooting that? And I think a good example here is something like airfares, right? They're still about 10% above, excuse me, below their February 2020 level, but they're more like 13% below where you would expect them to be if they had

just continued on their pre-pandemic trend. So that kind of tells you, you know, things normalize. That's an area where you might start to see some upward pressure come as an airfares. And on the flip side, hotel rates are running about 8% above where you'd expect them to be right now. And so that's the place where you might get a little bit of give back. And if you don't, then you potentially start to get a little bit of concern that this might stick a bit longer than you would have expected. Big picture. I mean, you know, we fixate on a few of these so-called reopening categories and used cars. We know the story there and rental cars. We know the story there with airplanes and hotels. We understand some of this. Also, you know, we talk a lot about certain goods related that relate to shipping and logistics, which we know is supply chains jammed. When you look at some of these measures like you do, the breadth of the inflation, general inflation momentum and so forth, what are you seeing right now? Is there a process happening where it appears to be broadening out?

and momentum is gathering steam, or is it something else? You know, let's, maybe we go back to this kind of the Fed's preferred measure of the PCE. The San Francisco Fed actually does a nice job keeping track of some of these dispersion measures and so on. And what you look at now is that about roughly, you know, if you sort of look at all the components in the PC, about 84, 85% of them currently are showing price gains. And, you know, that sounds like a pretty big number that most of the, most of the, most of, you know, the components are rising. But in fact, it's only a couple of percentage points

more than what we were seeing sort of, you know, pre-pendemic. So it doesn't, it's not clear to me that we've seen a big broadening out of price pressures. We've seen, as you mentioned, just really concentrated increases in pressures in some components. So I'm still very much in the camp that, you know, as we sort of get through the spring of 2022, we're essentially going to see a lot of slowdown, I think, in the core. And much more, I think, in the core PC, for example, like I wouldn't be surprised if by the middle of 22, we're talking about core PC being closer to around 2%,

whereas the core CPI potentially is still, you know, punching along at around 2.5%. And one thing I just want to mention is, you know, you talked a little bit about some of these macro stories with, you know, chip shortages and so on. You know, these are important, right? We like to have narratives to try to explain something. But one of the things with when you,

are really in the weeds of this inflation data is that as important as those narratives are to kind of understand the picture, most people don't really pay attention to the fact that a lot of price movements that you tend to see have nothing to do with a macro story or a micro story. It's literally just about the way the methodology works in the index. It's about the seasonality of the index. You know, it's about changes in the way we actually compute and construct the data. and it has less to do with, you know, these broader stories that we're trying to explain, we're trying to use to explain the inflation number. Sometimes it's just about understanding how this thing is built and sort of really getting into the weeds of, you know, understanding the

parts that sort of make up the sum. And that good example is motor vehicle insurance. This is an index that, you know, a lot of people don't pay attention to. Last month was down about two and a half percent, you know, a little bit less than a full tenth off the core CPI, which is a lot when the course around 0.3 only. And it has nothing to do with a big story. You know, insurers are not cutting your rates. It just has to do with the way the seasonality is working out this year for this particular index. And it's going to be a very similar story when the next point comes out. Insurance should be down around 2.5.3% again due just to the seasonal factor. So no big story. But if you understand the seasonality and you understand how this thing

is constructed, it gives you an edge in terms of forecasting this number. Just to play devil's advocate for a second. So when it comes to those macro stories that you just mentioned, one of the things that Joe and I have been discussing a lot over the past year or so is this idea of the bullwhip effect and that you end up seeing a massive amount of volatility in orders and stockpiling because of the uncertain environment. So you know, you get a shortage one month and then everyone ramps up their orders because they don't want to be caught short again. And suddenly they're oversupplied and you get these sort of intense price increases and decreases. So I guess my question is like, is that a risk to inflation actually proving transitory?

Is that something that could start to come into play? Yeah, that's a good point. It's possible. But I think, you know, where you would see that obviously would be a ramping up in, you know, in the inventory numbers for some of the, you know, the places where we're seeing, where we are seeing shortages now, where we are seeing orders pick up steam. You know, I mean, honestly, to some extent, to your point, we're kind of seeing this with used cars now. So wholesale prices have been coming off the last couple of months. And again, where you really want to look is on the

inventory side. So when you look at used cars, when you look at the wholesale piece of that, we're only about a couple of days below a normal level of inventory for wholesale used vehicles. And on the retail side is actually pretty similar as well. So, you know, we're catching up on the inventory side and getting to something that actually really resembles normality in the wholesale used vehicle market. And, you know, lo and behold, we're down about 3, 4% in wholesale prices over the last couple of months. So I think sort of to your point, I think we are starting to see some of that happen.

in some of these components. Hi, I'm PJ Vote. My podcast search engine has a new two-part series for you. Of all the new technologies coming out of AI, the most transformative one might be driverless cars. They're already on the road in 10 American cities, and they're quickly coming to more.

We tell the story of how we got here. The secret team at Google that spent 15 years building what might be the safest vehicle on the road, and we cover the fights brewing in blue cities, where unions and politicians are working to keep those cars off the streets. Listen to search engine wherever you get your podcasts.

So I'm thinking, you know, I kind of want to zoom out a little bit and talk about the relationship between some of your work and, you know, how investors use it. You've been on the sell side. You've been on the buy side. Now you have your own shop. As Tracy and I have talked about in past episodes, inflation gets people going. It gets consumers going emotionally, but it also gets traders and investors going and people have very strong views about the Fed and so forth. I'm curious about like receptiveness to your way of thinking. because you obviously clearly take this bottoms up approach where it look not only at individual categories, but individual category construction. How do like, you know, traders and investors who want to use this, are they receptive to it? Are they, do they get, you know, are they angry at the ideas like, oh, you know, this is all the Fed's money printing, et cetera, which is kind of seems to be the opposite of how you think about these questions. What do you talk a little bit more about your work and how investors use it?

I think by the most part, people are incredibly receptive to it. it. I mean, when I started doing this, I don't know of many shops or many individuals who were taking this kind of bottom approach and sort of doing a detailed analysis of the components and index construction and so on. And I think people are, especially periods like this, they want to understand what is moving to print. Is it a one-off print? Is there, you know, was there something driving it this month that could be more persistent? And how do I think about that for the following month? Because if you're an investor and you're, you know, in the tips market, or you're, you know, interested in the fixings, one month obviously influences everything.

And so you really want to understand what is what's going on kind of beneath the hood of the data. So I think people are incredibly receptive. And, you know, in terms of getting pushed back from folks who are like, hey, this is just the Fed's money printing. There's, there's always some element of that. But I find that those are the folks who are, you know, potentially removed from actually trading or managing money. You know, the folks who are managing the money, they, are into the weeds of it. And, you know, it's funny because as I mentioned when I was on the cell side, when I did this very few people did it. Now being on the buy side the last two years, I was a consumer of all the cell side research. So I got all the research from the banks on

inflation and how they forecast it and so on. And it's funny to me because now a lot of people take this bottom up approach on the sell side as well, some more so than others. But it's kind of the way everyone on the cell side is doing it now. Because I think there's a value add in understanding the weeds of, you know, what's driving shelter inflation and what's driving apparel prices and so on? Because it really does give you a window into where the kind of headline number is going. And importantly, is it going to stick or is it just, you know, kind of a one off? So in a bottom up approach, like the one you just described, what role, if any, does monetary policy actually play? And, you know, I'm thinking of that famous Milton

Friedman quote about inflation is always and everywhere a monetary phenomenon. on, like, is that incorporated anywhere in the type of work that you do or is it irrelevant? The way when you're doing this sort of approach, it is almost by design. It's a very short-term approach. You know, I'm not going to say it here and say, hey, I'm in a forecast inflation for the next five years doing this approach because it's just not designed to do anything like that. When you're doing something like this, it's much more looking at, let's say, the 12, next 12 to 18 months. And really, if you think about policy and the lags, you know, to some extent, you know, maybe it's impacting some of these components in that time frame, let's say especially housing. But in the 12 to 18 months, it doesn't play that big of a role. And honestly, if you even think about the way that these indexes work, the San Francisco Fed a couple of years ago had a great paper where they applied kind of the Phillips curve methodology to individual components of the PC.

and what they found was roughly 60% of these components are what they would term acyclical. And essentially, you know, policy can't really impact them. So stuff like medical care, for example, whatever you're doing with policy is probably not really going to impact physicians prices or, you know, hospital prices. And so 60% of the index just doesn't really react to policy. And even the 40% that does, you know, it's going to be a bit of time that's 12 to 18%. months, and this approach is pretty narrowly focused on kind of, you know, just that kind of window. So I would say that if it does, it's kind of hard to really, it's hard to really incorporate it into

this sort of a framework when you're thinking about forecasting. So we sort of establish that there isn't that much of a sort of cogent theory of inflation from sort of like pure macro standpoint. What is your pitch then? Is it just that you're going to help, like explain? what you, a little bit more about your pitch to potential clients to help them understand what's going on. Like, what is it that you say it's like, okay, you do at your new shop, appropriately enough called inflation insights. What is like the basic sales pitch of what you can

bring to the table? Sure. So, you know, for me, my target audience is mostly, you know, institutional clients, right? The folks who actually are trading tips and who are trading the fixings. So in that respect, the pitch is really sort of, I would say there's kind of three main elements. One is the actual forecast. You know, for me, luckily, I've been doing this long enough where I've got a reputation. I've got a track record in history that I can present to clients and say, look, what I'm trying to build here is the best in class forecast that you will get on the CPI's NSA index, which is what matters for tips.

And kind of here's my history of that. And that's the goal is to have that be the best in class moving forward. The second is just the detailed analysis, making sure that everyone understands what's going on. And the timeliness, I think, also matters quite a lot. So, you know, the stuff that I put out typically will be well in advance of anything you're going to get from the south side. And it will give you an opportunity, you know, if you agree with my view, for example, that it'll give you an opportunity to actually trade it in the market before the CPI comes out. Whereas right now, a lot of the cell side research, you know, it's coming out 48 hours before the number prints. And that's really not much of an edge.

But it's the detailed analysis, the timeliness. And then finally, you know, I would say I'm probably on the horn with the BLS, if not daily, you know, at least once a week. Even though I've been doing this for a long time, it is honestly just a constant kind of learning process. I mean, there's there's about 211 indicators that go into the CPI. there's over 7,000 basic item in area indexes that you could look at. And so it's just kind of a constant learning process. And it's, you know, for me, I've always had this luckily good rapport with the folks there who I think are incredibly helpful in terms of learning about the components and so on.

And that's sort of the kind of, I think, you know, the kind of knowledge you're not really going to be able to get most other places. So now I have to ask how specific you can actually get. when it comes to the inflation baskets. So this is a really weird question, but I went on like a massive tangent a couple weeks ago because there was a restaurant in North Carolina that a guy was quoted as saying

that he was spending $200 more per week in mayonnaise because of inflation. And so, of course, everyone started calculating like, well, how much mayonnaise is this restaurant actually buying based on CPI? And then I started going on Bloomberg and looking at the components in CPI and it turns out mayonnaise comes under the salad dressing and spreads basket. And so I

guess I'm just curious, like, how in the weeds do you go? And can you give me like a quick, a quick read on what's going on with with salad dressings? Yeah. So I don't know if I can pull up the salad dressing forecast just now. But, you know, I would stick when you get to kind of that level, you have to make choices, right? I mean, like I said, there's over 7,000 item area indexes. There's over 211 sort of broad components in the CPI. Most likely when you kind of get something like food, which has dozens of indexes,

you kind of have to make a choice in terms of how far you're going to drill down. So I might follow all of these, you know, I've got them in my spreadsheets and so on. But when it kind of comes to forecasting, you're probably going to want to stick, for example, with looking at the broader two. which is the food at home index, which kind of encompasses the entire grocery basket, and the food away from home with restaurant prices.

And there, within food at home, you know, you would, if you want to drill down, you would break it down into some of these components. So cereals, you know, the various types of meats, eggs, food, vegetables, and so on. But it doesn't mean that you're necessarily going to go in and, you know, forecast uncooked beef steaks, for example, right? I mean, you could, you could, but it would take you a month or longer to just come up with a simple forecast. I mean, I used to spend probably two days just doing the food forecast. So you have to kind of make some choices about the timeliness of your forecast and how

into the weeds you're going to be able to go in order to produce something that's actually, you know, actionable. So this reminds me of something I wanted to ask you. So when I was in my Mayo analysis adventure, one of the things I was trying to do because I couldn't find an inflation pickup in the official CPR. basket, but I tried to look at an Amazon tracking website to see if prices had gone up on Amazon. So I'm curious, do you ever look at alternate data in order to make your forecasts? Yeah, so there's a couple of things I do for certain components where I will look at, you know, non-BLS data sets to try to get a sense of what's going on. And one of those, for example,

is just airfares. Airfares is only worth, you know, a little bit less than 1% of the core, but it's basically been the bane of my existence in forecasting for the last, you know, 15 years because it's incredibly convoluted the way that it's done. But, you know, at the end of the day, what they're really pricing is, you know, they're going to the websites of, you know, Delta, America, and Southwest, so on. And they're pricing flights out. So you can try and come up with an index yourself where you just, you know, go onto these websites and try to say, hey, what's the flight, you know, from New York to LA going to cost me or New York to Miami or whatever. And so you can sort of look at those sorts of data to try to help you forecast the airfare

index. You can look at, you know, Black Book and J.D. Power and so on to get a sense of what use car prices might do. And then there's, of course, the Billion Prices Project. There, I think you just have to be a little bit careful because a lot of what they are capturing is much more, has much more to do with goods prices and much less. to do with services prices. But for goods prices, you know, that does a pretty decent job from time to time. So there are other things that you can certainly look at. GasBuddy, which is actually now being used directly in the CPI.

So they've gone from having, you know, 1,000 quotes on gasoline each month to having millions of quotes because essentially they've crowdsourced the data is something else that you can also look at. So there definitely are alternative data sets that you can try to work into your forecast. I just want to say, Tracy, you stole that question right. That was literally the next thing I was going to ask. Oh, I'm sorry. No, no, no, that was great. You asked it great, but I just continue to be perfect wavelength.

You know, I'm a bigger picture, or I guess sort of like medium term, you know, you present, as you said a little bit ago, when you look at some of the broader metrics, you don't necessarily see a sustained upward move in inflation, that there isn't necessarily this kind of momentum that even if rent were to go above. grow at a pace that's well above historical averages. They might be offset. What would make you worried or what would make you think, okay, this is going to be a type of elevated inflation that persists. And maybe monetary policy doesn't really affect inflation, at least in the medium term. But inflation could certainly affect monetary policy if the Fed gets spooked or so forth.

So I assume that's important information for investors. what would you be looking for, say, through the rest of this year or early next year, to say, oh, this is going to be higher and more persistent than I would have guessed. Yeah, so one of the things that's, you know, kind of come up in the last print or two that I'm going to be keeping a pretty close eye on going forward is this, you know, idea of whether some of the pickup we've seen recently in wages begins to pass through more persistently into the inflation data. And, you know, we saw this actually last month in the Food Away from Home Index.

there was a pretty big increase in what are called limited service restaurants, so fast food. And it was for that index was a huge, huge jump. And we know that wages are going up in leisure and hospitality, for example. And so the idea that some of this might be feeding through, for example, into, you know, hotel rates or limited service restaurants and things like that, those are areas where potentially you start to say, okay, you know, if we keep seeing wages move up at these sorts of rates, if this is what, if this really is that kind of a pass-through, then this is potentially something that is more persistent that will last into next year. And it's not going to be something, you know, sort of a one-off shock like let's say, you know, oil price shock or something of that nature.

You know, this is something that is potentially more persistent. And I have to say, I think even with something like used cars, we know they're starting to come off. I am a little bit wary of just kind of having a repeat of what we had, which is, you know, last summer we had a huge jump in used car prices. It completely tailed off. They declined throughout the fall and winter until we, you know, had another huge burst over the last several months. And whether that's, you know, mostly a function of sort of the demand side or the supplies, I'm definitely more on the supply side part of that story. But, you know, it's, we still have kind of to deal with this idea of the delta variant. And what's, what is that going to do to activity going forward? What is that going to do

for the demand for use cars and new vehicles, you know, later into this year and into next year? I don't know that anyone's got a good answer for that, but that is something that I think kind of remains an upward risk in the inflation story is. You know, we sort of see a repeat of some of these upward pressures from, you know, something like the Delta variant going forward. But more persistently, it would definitely be some of this wage pass through into some of these components that I mentioned earlier. So on a related note, is there, you know, in your very long career analyzing, inflation, is there any particular component that has just remained an absolute mystery to you and that is sort of like, I guess, immune to the bottom-up analytical approach? Like, something that really flummoxed you. Yeah, I think apparel. Apparel is one where in, you know, 15, 20 years

of doing this, I've literally just never found anything that works at forecasting apparel. Other than, And, you know, one of the approaches I mentioned earlier, which is just this naive approach of saying, okay, you know, looking at these things on an unadjusted basis, you get a sense for sort of seasonal patterns. And, you know, let's say apparel every March tends to decline by about two tens. You're going to be probably doing a decent job if you put in, you know, a drop of two tens because there's very little to go on when it comes to to apparel prices. You know, I've tried using everything from import prices to, you know, different data sets, retail sales, and so on. And there's just nothing that gives you a good lead into what apparel is doing. And that's been, you know, that's been another tough one to do.

Not as difficult as airfares. Airfares, at least you know what they're doing. You just can't replicate it exactly. But apparel's, yeah, apparel's one that has just really, you sort of just have to look at the patterns. And honestly, it's, it's, it's much more about kind of looking at these patterns and thinking about, you know, to what extent you're following the same trajectory as you did in the past for something like peril. So one of the sources of like constant controversy and, you know, CPI, inflation truthers and so forth, always like to talk about it is like the so-called hedonic

adjustments. And they're always like, oh, this is not, you know, we've all heard the conspiracy theories. Right now, for example, though, if you, if everyone is, you, complaining about the service they get at restaurants because of the so-called worker shortage. People are complaining about the service they get at hotels. And we know, for example, that hotels in some cases have degraded service or not picking up towels as often or whatever it is. Are these sorts of things captured? Does the BLS, I mean, as you say, you talk to them all the time and you're trying to learn their approach, are they trying to capture these types of things such that maybe the experience at a restaurant or a hotel isn't what it was in 2019?

No, so hedonic adjustments are just applied to goods and not to services. So, yeah, and so, you know, if the, I'm sure you kind of remember this whole, you know, wireless thing in March of 2017. Yeah, so things like that is where you see hedonics. Apparel is one where you see hedonics. And so really is just limited to goods. And also, when you think about hedonic adjustments for the most, you know, part. I think it's about only about four or five percent of the CPI is actually subject to those sorts of kind of quality adjustments. You know, other indexes are subject to other types of

adjustments, but they tend to be much smaller. So, for example, with rent, you know, rent, there's something called an age bias adjustment because you're, let's say you've got an apartment and you happen to make it into the BLS survey. And in January, they come to you and they say, you know, what do you pay for rent? You give them a number. they come back to you in July six months later. And of course, let's say you've got a year lease, your rent hasn't changed. But your apartment's six months older. And so they apply what's called an age bias adjustment to your apartment.

It doesn't really change, you know, very much. But those are the other sort of types of adjustments that the BLS will make. But for hedonics, it's just not a big fraction of the index that is really kind of getting that kind of a treatment. And it's really only, you know, it only really comes to light when you have these huge moves like you did with wire. a couple of years ago. Those sorts of quality adjustments, you know, they get a lot of press, but they don't really... Can you remind people what happened? I remember the wireless thing happening, but I don't remember what it was.

Can you just remind people? Yeah, I think we went to unlimited wireless plans in, I think it was February, March of 2017. So, you know, when we switched over from, let's say, you paid however much your bill was for a certain amount for your phone, when we went to unlimited plans and you had, you know, instead of having... you know, 10 gigs and you maybe you had 30 or whatever it was, they had to find a way to price that out. And that's where kind of the hedonic regressions came in was to say, you know, how much is this extra speed worth or how much is this extra memory worth? How much is this extra data plan worth to the consumer? And once they came up with those measurements, they applied

them and what it ended up leading to was about a 7% decline in the wireless index in a one month, which is a record decline. And it subtracted about almost close to a 10th, just a bit over a 10th, off of the monthly change in the core CPI. And that's a huge, huge number. Got a lot of attention. And so people started talking about hedonics again. And that's where, you know, the so-called sort of inflation truthers come in is like, well, this is just this arbitrary adjustment that they're making and so on. But, you know, this is the kind of stuff that goes on on a pretty regular basis. Typically, you just don't see those kinds of moves. But for the most, part, these things are pretty standard. And not just for, by the way, not just for the BLS, but almost for

every statistical agency that does a CPI. At the end of the day, the CPI is, you always want to price between, you know, one month and the prior month, the same good. And if it's changing in quality, you have to try to control for that quality. So these adjustments are happening, you know, Canadian CPI and Eurostat and so on. So it's a pretty sort of time-tested methodology that everyone, you know, uses in all kinds of consumer price indexes. So I just have one last question, and it's sort of big picture, but, you know, in the very beginning you talked about, okay, different regimes, different times, different relationships might work.

Phillips curve thinking sometimes seems to be robust, sometimes not so much. One of the things in the post-crisis period is people are asking, well, is this like a new regime? Like, is this, is the economy now just going to be fundamentally different? Maybe because of some sort of change to international trade. or something like that. Is that something that you're on the lookout for or thinking that maybe like even post virus, maybe we'll get something resembling normalization, but that something structurally might be a different economy than we had pre-crisis, thus forcing, you know, thus causing a sort of

different way to think about what might manifest inflation? Yes, I think let me preface this by saying, you know, for the most part, economists are really terrible at picking up like turning points and, you know, paradigm shifts and things of that nature. which is why there's such a large literature on how to forecast inflation. But yeah, I think one of the things that, you know, at least I'm on the lookout for it, and I think others are as well, is to think about the idea of how all this sort of disturbance and supply chains is potentially going to lead to, let's say, onshoreing. You know, we're talking about building more semiconductor factories here in the U.S.

We're talking about having sort of, you know, more of, you know, manufacturing in the U.S. And, you know, what does that mean for, for inflation going forward? So that's potentially a big paradigm shift that I think we need to be on the lookout for. But is that going to be a 612, 18 month thing? You know, I'm pretty skeptical of that. To me, that is a much broader, much sort of, you know, longer tenure type of topic to think about. And, you know, probably not something that you're really going to be able to capture kind of doing a bottom up type forecast. Omerer Sharif, thank you so much for coming out in Adlaught.

Really appreciate it. Thank you. appreciate it. Thanks, Omer. That was fantastic. Thanks, Omer. I really like, that was great. I really like talking to Omer. I feel like at least right now, it really feels like if you're not doing some sort of bottoms up analysis where you're actually looking at the component, there's probably like no hope to understanding what's going on with inflation. Yeah, totally, especially since so much of it seems to be driven by the reopening, like not just the reopening categories, but literally one or two. or maybe three reopening categories,

like a big chunk down to used car prices, airfares, and hotels, I think. Yeah. I also thought it was super interesting that, you know, it's like that he sort of pushed back a little bit about the so-called like stories we tell about even those categories. And so even though like, okay, we can talk about semiconductors or shipping containers, but that actually, per his view, you have to go even deeper

and just like really get to know index construction. and really no methodology and seasonality. And to actually sort of do it is not enough to just be able to like sort of like tell some like bigger stories about the categories that are really moving. Yeah, I am very curious about the seasonality portion of it. And I guess like if everything is so seasonal and predictable, why do people still get it wrong occasionally? I guess it goes back to what we started the episode talking about, which is this idea that, you know, despite decades. and decades of studying inflation, it does feel like economists certainly struggle to look at it as a whole. It's also interesting, by the way, I really liked your question about mayonnaise inflation or the salad dressing category.

But it is interesting that there were like these categories that he like, you know, he expressed sort of like confidence about his ability to make a forecast and then other runs. And I think he said like airfares. And you wouldn't necessarily think it with airfares because, again, it seems like the numbers are kind of transparent. or there's like dozens of websites that track airfares. But it's interesting that there are like these categories that like you just can't quite crack. I have a great book idea. So what if you went through the like 200 CPI components and like for each one kind of told a story of the industry and how prices are actually made and how the BLS incorporates them? Yeah, best seller.

New York time. No, I think it'd be interesting. I've read a book once that went through every ingredient of Twinkies. And that was fascinating. There's something like 100 ingredients in there. You could do the same for CPI. No, actually, unironic. What about a coffee table book?

Like each picture is sort of like a really glossy, beautiful photo of like mayonnaise or plane or something like that. And then a page on the left talking about how the prices derive. Yeah. Okay. Literary agents and publishers hit us up. We're ready to write it. Reach out. Let's leave it there.

All right. leave it there. This has been another episode of the All Thoughts podcast. I'm Tracy Allaway. You can follow me on Twitter at Tracy Allaway. And I'm Jill Weisandthal. You can follow me on Twitter at The stalwart. Follow our guest, Omer Sharif. His handle is at F-cast of the month. Follow our producer, Laura Carlson. She's at Laura M. Carlson. Follow the Bloomberg head of podcast, Francesca Levy, at Francesca Today. And check out all of our podcasts at Bloomberg under the handle at podcasts. Thanks for listening. I'm Francine Lacqua, an award-winning journalist, and I've got a new podcast, Leaders with

Francine Lacqua from Bloomberg Podcasts. I've interviewed everyone from Heads of State to fashion icons about the news of the moment. But I've always been curious who are these people as leaders. I don't think there's one right way to be a leader. Make decisions. A poor decision is always better than no decision. Listen to new episodes every other Monday. Follow leaders with Francine like why, wherever you get your podcasts.