Tim Graf (TG): This is Street Signals, a weekly conversation about markets and macro brought to you by State Street Global Markets, the Markets and Financing Division of State Street. I'm your host, Tim Graf, European Head of Macro Strategy. Each week, we bring you the latest insights and thought leadership from our award-winning suite of research, as well as the current thinking from our strategists, our traders, our business leaders, and a wide array of external experts in the markets.
If you listen to us and like what you're hearing, please do subscribe, leave us a good review, get in touch with us. It all helps us improve what we hope to bring to you. With that, here's what's on our minds this week.

The concept of greedflation, the notion that businesses took advantage of broader inflationary trends during the post-COVID period to raise prices more than their rises in costs, thereby fueling further inflation, it's getting a lot of attention again these days. Without naming names, a high-profile merger of US grocery store chains has brought the topic back into the headlines after it was also a widely discussed topic during that COVID era period of high inflation in the US and elsewhere.
But does the belief that retailers used the pandemic to their advantage hold water? Well, that's what we're going to talk about this week with Alberto Cavallo, one of our academic partners. Alberto has done a load of other work on inflation, but he has a new paper out on markups along the supply chain, where greedflation might be apparent. Alberto will be familiar to long time listeners and readers of our research, but for the new ones, he's the Thomas S. Murphy Professor of Business Administration at Harvard Business School and co-founder of PriceStats, a leading private source of real time inflation statistics for over 25 countries.
Hey, Alberto.
Alberto Cavallo (AC): Tim, how are you?
TG: I'm good. How have you been?
AC: Great. Great to talk to you again.

TG: You must be pretty busy talking about this paper.

AC: Yes. In fact, we put it out tentatively as a working paper or we're still working on it. Yeah. And I said, okay, let's just put some facts out there. So I put it on Twitter and it got a lot of attention.
TG: Yes.
AC: So now I presented it a few days ago at MIT and other people are asking us to present. So there's obviously a lot of interesting decision. And we have unique data, so it's absolutely.

TG: Absolutely. Twitter is how I got to see it actually, because I follow you, of course. Just to give background, the paper that Alberto is referring to is called Markups and Cost Pass Through Along the Supply Chain. And it's kind of an innocuous title if you don't read the news and you don't see some of the headlines particularly around the merger of a couple of food retailers. There's been a lot of discussion about price gouging. And so having Alberto on for this week, I thought was super timely because this is still a paper that is a working paper, a work in progress, if you will. But it is getting this attention. So Alberto, I wanted to see, you mentioned you posted on Twitter about the paper. You actually posted some of the key findings of the paper. I wanted to see if just kind of you can give us the bullet points at the beginning and then we'll get to the details.

AC: Absolutely. So the paper looks at how markups change over time along the supply chain. And like you said, this is a very topical issue because there's been a lot of interesting understanding what is driving the surge of inflation in recent years. And many people have pointed to markups as potentially being one of the reasons. Let me tell you the findings. We find actually that total markups, when you look at production costs of non-durable consumer goods, and you compare that to the retail price, the markup has been remarkably stable. Even though prices surged in 2022, we do not find any evidence of markups arising. And so this goes against these claims of greedflation, which is loosely defined in the media. Some people define it one way, others define it another way. But mostly, it's the idea that prices have risen more than can be justified by the increasing costs. And what we do find is essentially that prices have mostly responded to the rising costs and with a delay. And certainly, markups in our data have not increased/ So that's one of the main findings. And I think the one that attracted a lot of attention when I posted it online, because normally people discuss about markups, but they don't actually have the data to estimate them directly. And here we have one of the largest global manufacturers of consumer goods. We can see what they did with their costs, with their prices, and we can compare also with the markups at the retail level. And we don't find evidence for some of these claims.

The other finding that I think is very interesting in the paper is catching a lot of attention is why does this happen? We do show in the paper that it's important to consider a whole supply chain. If we were just focusing on retailers, the story would have been different. We actually did see a rise in markups for the retailers in 2022, although they came back down. And you would have had a very different story if you just look at the manufacturer of these goods. You really need to think of the whole supply chain when you're having these discussions about profitability, markups, and creating stories of how that might have mattered for inflation.
And thirdly, I think is important for people to understand, is that it seems to be a lot related to the availability of information for the manufacturers and retailers. The manufacturer is reacting in ways that suggest they know more about these cost shocks. They know what is happening. They know things. We have results in the paper about how they adjust to quality. These are all things they know much better than the retailer does. So in the context of a cost shock, the manufacturer ends up reacting sooner to these cost shocks and passing that on. The retailers are clearly less aware of the details, so they may be reacting more with, you know, category changes rather than product-specific changes. So it's a very enlightening kind of work in terms of helping us understand how the supply chain can transmit cost shocks into final prices. And eventually things like inflation.

TG: Before we get to, I want to get to how we talk about markups, how you determine a markup and some of the methodology in the paper, the very specific methodology and kind of, I think some of the confusion about this topic comes from, well, especially some of the confusion I saw in the response to this paper, first and foremost, I think didn't quite understand the data that you are able to produce with PriceStats and that you have access to. And so I want to talk about that.
But I think one of the other misconceptions people have about greedflation or perhaps have about greedflation is, first of all, it's not to say prices didn't rise, of course they did, but that a lot of the gouging, if it was to be had, may have happened, not so much at the retailer level, but at the producer level, at the manufacturing level.

Would you say it's fair to say that if there was greedflation to be had, it was much more likely to be done, based on what you just said, at the manufacturer level than at the retailer level? Is that correct?

AC: No, it's actually the other way around. We found evidence of rising markups at the retail level rather than the manufacturer level. Just to be clear, these can change and we look at different countries. There's a lot of heterogeneity on that split, and I think it depends a lot on the availability of information about the shocks and the movement of costs that each one of these parts of the supply chain may have.
But you're raising something very important, which is, which we emphasize in the paper, that you really need to understand the full supply chain. Just focusing on one side of it, we could come up with very different stories. And that often happens in discussions in the media. People say, oh, I know one retailer who did this, perhaps a manufacturer that seems to suggest that, and that creates a lot of confusion. So we really wanted to focus on the whole supply chain in this paper. And you're also absolutely right. One of the problems is that people don't really define very well what is a markup or exactly what they're talking about. And they mix it up with profits, for example.

TG: Yes.
AC: Certainly, there were a lot of companies, and you can see that in the public reports from these companies. There are a lot of companies made a lot of profits. But that could simply mean that they're just selling a lot more units than before and making the same markup they were making per unit in the past. We find evidence consistent with that. What we do not find is that the firms are selling goods or higher markups. And by markups, I mean a difference between the price of that unit and the cost to produce that unit. Those have remained very stable. How can you explain this, then? How are they making more profits? Obviously, the demand shock, the stimulus can help on this front. It can allow them to sell a lot more products than before, and that increases their profits. So I think it's important to distinguish that in exactly what will be my markups and differentiate it from profits.

TG: Well, one of the reasons I was so keen to do the podcast as soon as I saw this paper from you is I've been trying to think about how to discuss this topic, because I think there are a lot of misconceptions around it insofar as, to your point, you look at gross margins, you look at net margins of particularly things like, I think the sector level category is something like consumer staple, distribution and retailing. I think that's the S&P category for these companies. You look at their margins and they're very, very stable, but their profits are quite strong. As you mentioned, and that I think is one key misconception people have, is that the margins themselves are very, very stable, but these companies did raise prices. There was some degree of marking up, but it's not so much that they are excessively doing. So they're really just kind of, when it comes to net margins, it's kind of keeping pace with trends that were in place prior to the pandemic.

AC: Yes, or another way to say it is, yes, they are making profits, but that's not because they raised the markup itself on the individual products. It's, in fact, they're raising their prices because the costs have gone up. And we do not find any evidence that they're raising their costs, sorry, their prices more than their costs. So in that sense, the greed inflation story does not pan out in this day.
There's an important caveat that you probably saw many people responding to in Twitter. We're essentially using one good, we're using goods, a lot of goods, but it's only one manufacturer.
TG: Yes.

AC: And some people say it would have been possible or preferable to have more. Of course it would. Now my answer is I would want to have as many companies as I can. But since data is proprietary and hard to find, we had to focus on one large global manufacturer, which is the one that wanted to collaborate with us. But this is, by the way, much better than what anybody else has because we don't even have one. And they tend to estimate markups with techniques that require a lot of assumptions about their demand, the supply curves and things like that. And the other very reassuring thing is that this is a very big company. I cannot reveal the name for confidentiality reasons, but for decades it has been ranked among the top global manufacturers of non-durable consumer goods. So, if there is one firm that we might expect to be representative of consumer goods on a global scale, this company would be on the list. And so, from that perspective, I am quite confident about the generalization or representative of some of the findings we have.

TG: Yeah. That was one of the more silly critiques I saw of the paper, just kind of off hand, was saying, oh, well, it's not replicable. It's like, okay, well, you go out and you build a relationship like that, a confidential proprietary relationship with the manufacturer, and then build a huge business that tracks retail prices across millions and millions of goods. You go do that and then you can go replicate the work.
AC: But I do take it that it's a valid question. Oftentimes when we see results from research, academic research, we wonder how generalizable this is. And I am confident that we are looking at a very important company. And that tells me something valuable about the industry as a whole. I will mention, of course, that we are also talking about manufacturing. Is this the same in other industries? We still don't know. And I would hope other people would come in and provide some evidence for that.

TG: I do want to, there was some other questions that I did think were valuable to address about the paper. And I think you've addressed one or two of them already. But just to actually take a step back, could you walk through the data you got from the manufacturer and then the data you produce via PriceStats in terms of seeing how retailers might have marked things up in response to rising cost pressures?

Can you just kind of very quickly give us the description of how do you define a markup and how do you determine it using the data that you have to hand? Just give me some of that methodology as background.

AC: Yeah, so the key to the paper, to understand the paper, is we actually have direct data on the manufacturing cost for products. We have the price that the manufacturer charges the retailer. So think of it that as the wholesale price. And we can actually link it to the price of the product at the retail level. We do this by combining two data sets. Part of the data comes from the manufacturer, and parts of it, the retail side, comes from the information we gather through PriceStats. PriceStats, in the process of generating the inflation indices, collects data for products from thousands of companies. But certainly, these goods from this manufacturer can be found on many of these retailers. And I'm combining the data set here. So it's really unique on that perspective. Compared to other papers in the literature and other attempts to measure markups, the advantage is we can actually measure it with actual data, rather than having to estimate it.

And we can also, since we have these three pieces, the cost, the wholesale price, and the retail price, we can actually compare the markup for the manufacturer with the markup for the retailer. How is the markup defined? That's the difference between the price and the cost. So for the manufacturer, it's the wholesale price minus the cost expressed as a share of the price. Okay. It's as a percentage. When we do that for the retailer, we're actually doing the same thing. We look at the retail price that they charge. And here we compare it to the cost to them, which is the wholesale price. That's the cost of replenishing the item, if you will, the replacement cost. So that's essentially how this is built. So it's a very transparent way of calculating this difference between price and the cost to each one of these firms.

TG: So just very quickly, I'm guessing for this episode, we might be getting a lot of first time listeners because I have a feeling this is a podcast that might get a little bit of attention, similar to the way the paper has gotten attention online. For those who are listening for the first time, PriceStats is the company that Alberto co-founded who track prices in real time by using web scraping techniques across thousands of retailers in a couple of dozen different countries and then producing indices of price that is best we can map to consumer price indices. And then we track that through time and have a daily inflation measure.

So, Alberto, with that in mind, the question I have was, given you're using a large manufacturer with multiple product lines and then comparing it to retailer prices, is there any sort of, is it sort of like multi-channel retailer? Is there any specific type of retailer you had to use in the paper?

AC: These are large companies. When PriceStats collects data from these retailers, it usually wants to get at the largest retailers that have a presence online and offline. And if you look at all my research through the years, I have a work that shows that the prices we get online are similar to the offline prices. So many of the things I've seen on Twitter, for example, in reactions to the paper, are things that I can answer easily based on the past research that we have done.

And to be clear, I had always been able to see through PriceStats data, what happens to the prices of many of these goods. What was less clear was what was happening with costs. And in this particular paper, we are able to do that, which is why we can actually measure the markups and say something about it.

TG: Yeah. I wanted to get into some of the questions that were raised that we haven't talked about yet. I mean, one of the things that it's not private knowledge. Actually, companies themselves have talked about a strategy during the pandemic of pursuing price over volume. In other words, being willing to sacrifice a little bit of volume to get that price rise and in a way, it doesn't sound great, but in a way, it's taking advantage of an inflationary environment.

Again, that's not private or non-public information. I mean, how would you address that in so far as that's a public statement by, admittedly, I think more on the manufacturing side. Does that kind of comport with what your research has found?

AC: Yeah, that would be consistent with the economy experiencing only a supply side shock. Yes. That would raise prices and reduce the volume that we actually end up seeing. The pandemic was a combination of both a supply shock but also a demand shock. So we end up seeing the cost of production rising and prices rising by the same amount, and volumes have not, in the end, significantly changed. They actually recover a lot, particularly in 2021, 2022, as the economy started to recover.
I think what most people have in mind, though, when they talk about price gouging, for example, which is closely related to greedflation but not the same, is this idea that perhaps some manufacturers were taking advantage of the situation that there's suddenly this pandemic, this shock, the costs have increased, and they decided to pass it on very quickly. Usually, in these discussions, the unfair part comes into the story because there's a natural disaster, something like a pandemic, and suddenly the people who are demanding the goods are more inelastic. They're just willing to accept these higher prices, and that seems unfair that the manufacturer is taking advantage of this change in the elasticity.

But if that were the story, we would have found the evidence of the markups rising per good, because the markups are directly linked to the inelasticity of demand, or if you will, the insensitivity to the price that demanders would have. And that's not what we found. I actually got very concerned when I heard in the US a lot of discussion about price gouging, because it reminds me of some papers I wrote a few years back using data from Argentina. It's a very common reaction in Argentina. We see it all the time when you see inflation. People argue that the retailers or the manufacturers are taking advantage. They are the reason for this. And that leads to a series of price control measures. I've written papers that show that those policies always fail one way or another. They don't have much impact on inflation, and they in fact tend to be quite distortive on many dimensions.

But I think it's a slightly different story in itself, the price gouging story from the greedflation story. And I wanted to clarify that. The price gouging should be more, at least in principle, focused on there's a natural disaster. Suddenly, the people demanding the goods are more inelastic, and it's unfair for, even though it may be perfectly logical from an economic point of view, it's unfair for the companies to increase their prices at the moment in time. That may have been relevant in the first few months of the pandemic. The story that I'm trying to show evidence for or against, I would say, is more the story as the economy started to recover, is the reason we saw a sudden surge in prices of many things due to markups are rising. And that's where we don't find any areas of that.

TG: You talked about the elasticity of products. And this was another question that was raised. And I think the belief was that greedflation might have hit staples where the elasticity of demand is relatively inelastic, one would think, versus non-staples where that elasticity is perhaps higher.

Could you clarify that? And were there any dynamics that you saw within the data that either support or refute that notion, that staples might have seen a little bit more greedflation than non-staples?

AC: And that's a great point. In fact, I have another paper that looks at supermarket and grocery goods and in fact shows that more basic varieties of goods that you can think of, staples or cheaper varieties, they ended up having more inflation than premium ones. So in this paper, we wanted to sort of, is that related to markups or not? And to how quickly the manufacturer is passing on the changes. And we do have in the paper some interesting results that, in fact, we can distinguish the quality of the goods. We can't tell if they're staples or not. And in that sense, there's not much variation there.

But we can look at quality and we do find that on average, the markups for the manufacturer rise with quality. But we found the opposite thing with the retailer. So overall, if you look at the overall supply chain, the markups do not really change with quality that much. And perhaps more important for the question, the pass-through from the cost to the retail price does not seem to be greatly affected by quality, if you will, proxy for being a staple or not.

I will say, though, it's still possible that staples face more inflation for other reasons. But this paper shows it's not due to a change in the markup or potentially the sensitivity, if you will, of the cost shock becoming higher. Perhaps the staples are more exposed to commodities and they've suffered a lot from positive cost shocks in this period. Or as I explained in the other paper, I think there's also a phenomenon of demand playing a role there that when inflation rises, people switch to staples or to cheaper varieties, and that actually raises relative demand for these goods and ends up creating more inflation. But we do not find here any evidence that the markups played an important role in that.

TG: Were you able to capture any regional variations of retailers, say, within the US? This was a question again that was raised online that I thought was interesting because it does seem as though maybe the coastal cities, you could have seen some of these markup dynamics, maybe a little bit more in play versus in between the coast.

But were you able, first of all, to look at that and did you notice anything if you were?

AC: No, the data is at the country level, so we cannot. I will say though, I suspect we are likely to find more interesting differences across sectors than necessarily geographically because many of these consumer goods companies and even the retailers have uniform pricing policies within countries. So you wouldn't find much variation in this type of goods, in the prices that you would observe in one place versus the other. But you may still, I think, find a lot of different behaviors, whether we're talking about consumer goods or we're talking about services. I think that's where more research is needed to understand what is going on, and potentially a richer avenue for taking this forward.

TG: Well, you've created a lot of work for yourself, it sounds like, Alberto.
AC: Yeah, although it's hard to get the cost data in these other areas. So I'm not sure it might be so feasible.
TG: Your interns can focus on that, or your grad students, I suppose.
Listen, Alberto, I wanted to, while I've got your time, you know, I wanted to shift gears, and it's never a good podcast without talking to you about current inflation trends, because as we've discussed and introduced PriceStats or reintroduced it more accurately, we've talked about it a lot on the podcast, but I wanted to get from you the current message. And let's start with the US, because we have a Fed meeting that will actually be done by the time this podcast goes out. And the question will be whether they have cut rates 25 or 50 basis points. We will know that by the time people hear this episode. So let's take it as given that they will have at least cut 25 basis points. Thinking about how the Fed is talking about policy right now, it's pretty clear that inflation has taken a backseat to the risks in the labor market. Inflation expectations are pretty stable. We're getting back towards target. It's happening slowly, but then the Fed has said it was going to happen slowly, and we're not out of line with their inflation forecast for this year at all.

Can you talk a little bit about whether that is the right assessment the Fed has made to focus more on the labor market with inflation, maybe a bit of a dead issue now in the US.?

AC: Yes. In fact, our daily PriceStats measures, inflation measures, have been suggesting that inflation was falling and the target would be met. So the last time I publicly talked about this was in May at the State Street Research Conference. I was emphasizing that our data suggested that we were on a very good trajectory in terms of inflation. I think the Fed has internalized that now. If you look at what was CPI excluding shelter, it has been back to target for a very long time. So if anything, I was starting to get concerned that the Fed was taking too long to lower rates, but they seem to be on a good trajectory now, and I think it makes a lot of sense for them to do that.

TG: You mentioned shelter, and that is, I was traveling this week, but that was the clear headline that stood out as still being sticky. We do have a household equipment and furnishing series that also shows stronger than what is seasonally normal for that sector.

Is that really the only pocket that worries you that might argue potentially against a rapid normalization of policy, or are there any other pockets at a sector level within the data that might still cause some concern or is it otherwise just a pretty good story?

AC: No, it's just shelter. Shelter inflation remains persistent, but if you exclude it from CPI, like I said, the US has been back to target for many months. Like I argued in the research conference in May, I don't think keeping the Funds rate high is going to help bring that shelter inflation down much faster. In fact, it could be making things worse because it's making people less willing to buy houses and shifting to rents and keeping those rents high, which are in fact what we measure into the CPI.
But I'll tell you another reason I'm actually not that worried about shelter remaining persistent if we're talking about the target, which is, if you actually look at our daily indicators, in the last few weeks, and this excludes shelter, and also you can think of it as everything except shelter, they're actually showing another significant slowdown. It's still too early to tell how worrisome this is. But if anything, it will compensate for shelter remaining high and allowing the CPI to reach the target even sooner. So if I look at things that are not shelter, I'm actually getting worried that we are taking too long and the inflation may fall too much. So it reinforces the message I was giving you before, that I think the Fed is on the right track.

TG: Yeah. What about Europe here? Because they have had maybe some reason to believe that inflation is normalizing a little bit more slowly. And we had the ECB meeting yesterday that they cut rates, but they're still not committing to further action, at least further aggressive action. There is no 25 versus 50 debate in terms of the policy size moves in Europe.

Do you think that caution that they've expressed, maybe not very recently because they've just cut rates and that's the second rate cut, but they have been more reluctant, I think, to embrace easier policy than the US has. Do you think that caution is still warranted or is it now a similar story to the US?

AC: It was certainly warranted a few months ago. So I pointed out in May that the inflation metrics in our data suggested more persistence in the Eurozone and the UK series. And at that point in time, Europe seemed to be starting sooner than the Fed, and there was this uncertainty about how quickly they should move. So I think it made sense for them to be cautious back then. If you look at our data today, actually things have improved and still a bit volatile, but there's less concern now. Europe throughout this crisis seemed to be behaving always with a lack relative to the US. So just as the concerns about the US on inflation diminished, and now I think those are diminishing in the Eurozone, and they'll be more confident in the next few months.

TG: Well, last question. You've been very generous with your time, but one of the things that also stood out from your conference presentation given back in May were your thoughts on Japan. And Japan, of course, at that point had hiked rates once. They have since hiked rates again. But your take in between those two moves was that they maybe risked doing a little bit too much, even with rates as low as they are, and that the BOJ didn't really need to think about tightening further.

Would you still say that about Japan?

AC: Yes, I don't think they need to tighten much. And I know this may sound against consensus, but the PriceStats trend is very stable now at 2 percent. And my concerns in April were, if anything, that the inflation rate and the trend was actually falling, it has come down and it remains at that 2 percent. I worry more about the negatives of tightening. The overall inflation picture looks quite healthy to me, given particularly the history of a very low inflation in Japan. So if I were sitting in the position of someone at the Bank of Japan, I wouldn't be too concerned about this, you know, our CPI level being higher than the target right now because the PriceStats data in itself is showing a lot of stability.

And they certainly don't want to make the mistake that they've made many times in the past of raising rates and, you know, going back on their promise to keep inflation a target. I think it could become negative for them. If you look at the annual CPI, the annual rate of the CPI, it is still high, but mostly because of some base effects. The issue is not what has been happening in the last few months. It's rather because back in October of 23, prices rose a lot that we're getting above 2 percent in the annual rate. So if I look at the current levels and I expect the trend to continue, the CPI, the annual index, annual rate will be very close to 2 percent in just a couple of months. So I wouldn't try to rush that by raising rates too much. I think the risks are higher than the benefits for the Bank of Japan right now.

TG: Yeah, you've tried to get target inflation to 2 percent there for 30 years and now you have it. Don't mess it up, I guess, right?

AC: Exactly. That's the message I would give.
TG: Alberto, thank you so much for this. It's been great to kind of take a look around the world at the major economies as well as to talk through this paper. And again, I can't recommend it highly enough. And it really is one of the more timely pieces of research that I can remember. Just given the headlines, you've alluded to a lot of them, whether it is the discussion of greedflation. The notion of price controls you brought up with respect to Argentina, I think is interesting, just given there has been talk about that as you alluded to in the US. So such timely work.
The paper is called Markups and Costs Passed Through Along the Supply Chain from Alberto and his co-authors as well. I don't want to leave them out. It was a team effort. But Alberto, as always, thank you so much for talking to us.

AC: Thank you, Tim. Great to be here.
TG: Thanks for listening to this week's edition of Street Signals from the research team at State Street Global Markets. This podcast and all of our research can be found at our web portal Insights. There, you'll be able to find all of our latest thinking on macroeconomics and markets where we leverage our deep experience in research on investor behavior, inflation, risk and media sentiment, all of which goes into building an award-winning strategy product. If you're a client of State Street, hit us up there at globalmarkets.statestreet.com. And again, if you like what you've heard, subscribe and leave a review.
We'll see you next time.