Alberto Cavallo: Hello. A pleasure to be back with you. And trying to answer this question. Speaking of questions, as Eric was saying, that we've been asked a lot in the past year or so. As you can imagine, the question of whether inflation has broken lower or not has been in the minds of investors, policy makers and academics as well. So if you look just casually at some data, the question is very pertinent because we have seen inflation falling dramatically in most of the countries we monitor. We monitor currently daily inflation in 25 countries. What you're looking at now is the annual inflation rate of our World Inflation Index, which aggregates the data from all the countries where we have daily indicators. And you can clearly see since mid 2022, the annual inflation rate on a global scale has fallen and it does seem to be stabilizing now, but at a level that is higher than before, that little uptick that you see there towards the end is greatly influenced by the US. By the way, the print on the US just came out 20 minutes ago, a little bit higher, by the way, I should think the organizers for making this presentation 20 minutes after that's a gave me a lot of time to make sure everything is consistent. Good news. It is very consistent. We have been detecting that uptick. So what I'm going to try to do today is understand some of these dynamics in the past two years using the research and the data and the research we have produced for the past year or so.

So let me just say this is something you see in the US, but we've also seen it in many countries. This chart actually shows you the range of annual inflation rates we have been measuring online in the past 12 years since we started our work with with State Street and Price stats. And I showed this chart last year. It shows you the blue bars, show you the range, and the red dots show you where we are right now with annual inflation last year, everyone was on the right, far right. We were at peak levels of annual inflation. Those numbers have come down nearly everywhere. But I will point out most of these countries are still above their average levels, which are denoted by those little red bars that you see there. So this has an inflation dynamics that requires us to understand not just what happened, but also understand where we might be going. So I'm going to split the question of whether inflation dynamics are showing that we have broken it or not into three parts. In fact, based on three different papers I've worked on in the past three years since the pandemic started, the first one tries to understand a little bit the process of this inflation. Why is this happening? So I'm going to give you a quick update of something you may have seen me present in the past based on the ability we have of monitoring stockouts and supply disruptions online with that paper on stockouts to try to answer why has have we seen this disinflation without recession, this immaculate disinflation that some economists call it? Second, I want to think about turning points or the breaking.

When did that actually happen? And I'm going to try to answer the question of how do the new trends compare, by the way, to what we had before Covid. And perhaps for those of you who are very tuned in to the discussion today and just saw the numbers from the US, are we observing or detecting a new turning point right now? And then finally, I'm going to talk about a recent paper I wrote to try to understand why inflation rose so quickly in this particular crisis, why central banks seem to have missed this, and also what that tells us about the slowdown and the dynamics we may see in the future. So let me address each one of these in in steps. But I'm going to go to Slido first and ask you this question. I want you to tell me in just very few words why you think we have seen this inflation without a recession in this is not what most economists expected, of course, because as you start increasing rates, you create a decline in demand that has to produce some sort of a slowdown in output and some increase in unemployment.

And this was the discussion about hard or soft landing we were having last year. I should say. Most economists expected some degree of recession turning up. So let me see. What are your your thinking? So soft landing is exactly what we're trying to detect here. So most of you seem to think it's an improvement in supply. Fiscal fiscal would definitely have an impact if it helps monetary policy as well, not just on the decrease on demand, but fiscal policy that helps on the improvement in supply, as you're pointing out. Good. We have lots of issues there. Helicopter money that's interesting, robust labor market. But the big the big the big answer that you see there is the improvement in supply. And you're absolutely right. I may have given that away when I introduced the paper. But let's go back to my slides. I want to show you some data that backs this up. This is an updated graph from a paper I started writing when the pandemic started back in in early 2020 to try to see if we could leverage price stability to see not just prices, but the availability of goods online. And we created this stockout indicators that had risen dramatically at the beginning. And initially it was very obvious we could see the stock outs on the websites. Eventually the stock outs disappeared, but we had just seen a collapse in the number of available goods for sale compared to pre-pandemic levels.

And that's what this chart is showing you. The the graph for the US, a big increase in the early days of the pandemic reaching nearly 45% and then a gradual improvement. Now, a year ago, we were not in necessarily a very good shape. What had happened was that we had seen the increase in stock outs once again stemming from supply disruptions that were rippling through the economy and visible in goods. In the case of the US, mostly, many of them related to the war in Ukraine. But what you see towards the end of this sample is a dramatic improvement that started towards the middle of last year and has continued and we are now back at pre-pandemic levels. There is no evidence if you look at stock outs and supply disruptions in the goods we monitor online that there's any this disruption, let's say, between demand and supply. And this is an important part of the answer. As I mentioned during the bullet points, this improvement in supply is an important driver of the recent disinflation process. It's not magic, it's not immaculate in some way. It's just simply supply and demand. Now, it does tell this graph does tells us something about the future. Interesting, which is where do we go from here? There's no more room for improvements on these dimensions.

So we have to ask, how can we keep this going and bring inflation further down if this is running out of steam? So I'm going to try to address that in the next slides. Let's go to another question that speaks more about the turning point. Then we know there's this inflation. The question becomes when does this actually happen? So I'm going to make it a little bit harder for you. I'm going to ask you about core in the US. When do you think it happened? I'm giving you two dates. November of 22? Not yet. Or May. Very good. So. I see you've been looking at. Oh, good. May 23rd. Okay. So not not that long ago. That seems to be the prevailing answer. You're actually wrong. Let me show you. November 22nd was the right answer. Okay. I was hoping you would make a mistake. So I waited until that actually happened in the slides. So let me show you how we actually can measure this. Okay. This was record. Let's move to my slides again. We've always emphasized that our indices are using online data. There are some services we cannot cover, so getting the level identical to the CPI was never a goal. But we are very good at detecting changes in inflation trends, and we've always shown this graphically. In fact, these two graphs come from a paper we just published with my colleagues at State Street and Will Kinlow, Megan Sassone and Dave Turkington called inflation hedging using a dynamic approach.

And it just illustrates two historical moments in time when our indices started turning before the CPI did. On the left, you see the global financial crisis that I've shown in the past in Exhibit four there on the right, you see what happened at the beginning of the pandemic when our prices started rising before the CPI. Now we've always emphasized this graphically. What I want and I introduced last year was a methodology to detect these turning points statistically. So let's start thinking about how that would work. If we think of the US index right now, what I'm plotting on the left is the price that's aggregate index. In the last two years, three years. You can see there are clearly periods where the trend of inflation seems to be changing. And I've argued in the past and I will do it again today, that if you're talking about the trend of inflation, it's much better to look at the price index than to rely on annual or monthly inflation rates. Why? Well, as at least they're on the right, the slope of a price index is essentially that trend of the price level and it can last not a week, it can last a month, it can last years. So if you constrain yourself to annual and monthly rates, you would be missing much of the point.

The other advantage of doing this is that it's not affected by problems that we normally associate with the use of annual or monthly inflation rates, for example, things like base effects where we see numbers on the annual rate changing not because of what is happening right now, but because of what happened 12 months ago or price level jumps that often happen in some countries, for example, that reduce or eliminate price caps or impose price caps on some some sectors. And then another big problem that I see with the use of data of annual rates or is the fact that many people take monthly rates and then analyze them or quarterly rates and they annualized them. But that's not really the trend. The trend is that line that you see there on the price index. And that's what I'm going to try to measure today. But I'm going to do it not just at an aggregate aggregate level. I can do this at a very disaggregated level and then build an indicator of what I would call a turning point, a moment in time when the sufficient number of sectors and particular weights of the CPI basket have shown a change in these trends. Okay. So I'm not going to go into the details If you want to see how this is actually measured, you can read the paper, but only academics do that. And that's a very small share of the population.

So I'm going to focus on what's really important, which is the actual trends. Now, if you look at the aggregate index, what you would notice is that it had a turn. Let me go. Actually, can we go back one slide? I want to emphasize this. If you look at when those structural trends are detected, there's a slowdown that is pretty substantial in June of 22. Okay. That's when the the line becomes much flatter and then it goes back up again in January of 23. Now, if you look under the hood, let's say, and try to understand what is driving this aggregate effect, what you would notice is that this is mostly driven by energy. Many of these aggregate changes we see are almost entirely driven by energy. On the left, you can see what happened with transportation in the US. Big change, right? In June 25th of 22, a big dramatic change in the slope that was positive and that started being negative. If you focus on other headline sectors like food, there was also a change, but it happened actually much later towards September of that year. If you go to core, you would see also tremendous differences. For example, on the left you can see household and furnishing a very important durable goods sector for core indices that we can monitor online. That actually had an early break shortly after the Fed started raising rates. But the magnitude of the change was not particularly large.

And then we had things like electronics that. Haven't experienced any break at all. There are some sales that you see there towards around the holidays, but no change in the pre and post trend. So I'm going to do this not just at this level, but I'm going to go deeper and get the lowest level of aggregation I can for which we have weights of the CPI basket. I'm going to repeat the analysis so they take the dates when these breaks happen and classify these into positives and negatives, and then I'm going to add them up to produce an indicator of a turning point. This is what I'm showing you here. I'm actually fixing the start date in January 22nd. The worst time of the crisis. And then these two lines show you the shares of the CPI basket weights that experienced positive and negative trends over time. So it's accumulating. You can see that the the green line, that's the negative slope change is the slowdowns. Those weights were increasing quite gradually, but quite dramatically as well. Already by September, October of 22. We had seen more than half of the basket experiencing that slowdown. That's what I would call a turning dramatic turning point in the case of headline inflation. By the way, if we keep on going today, 83% of those sectors are lower than before. So we're clearly broken in terms of the trends, the inflation for for headline.

Now we can repeat the analysis for all countries in our sample. So I'm going to ask you, who do you think turned first? Of all the countries we monitor. Well, actually, I just put I just gave you four options. China, US, Brazil or Germany. All right. So let me stop there, because you're right. China was actually the first country we detected a turnaround. Let's go back to my slides. The ranking is actually perfect. China was first, Brazil was second. The US was third in this ranking. Actually, let's go back one. Sorry. There we go. So what I'm doing here is I'm repeating these charts showing the share the weights of the basket that had had slowdowns for all the countries. It's hard to see, obviously in that chart, but it's meant to illustrate how much difference in timings there are. There's a lot of heterogeneity here. Some countries turned first, others turned over a year later. And what are the countries that are turning first? That's actually China. We had a dramatic slowdown in the Chinese indices. We produced currently a food and a supermarket index. The turning points for those countries was already by March of 2022. The US looks pretty good on that on that table, particularly if we compare it to some countries in Europe, which has been laggards and taken much longer. They have shown similar dynamics but taken much longer. Although the latest results suggest that countries like the UK, France and Italy have finally turned here as well.

So there's a clear indication in headline. And also if you read the paper in core that we have finally broken inflation, that is good news now. The latest numbers in the US may raise concerns about whether we are again experiencing another turning point. So I'm going to shift a bit the focus and instead of looking at what happened since January of 2022, I'm going to produce something that is more useful for real time analysis. I'm going to use a rolling window and just focus on the last 12 months. So that will tell us something about what's happening with the annual inflation rate By definition. And what I'm showing you here is essentially the same indicators, the share of weights that have had positive and negative breaks. And when those two lines cross, it means we're entering a new inflation regime. So clearly there were two periods quite distinct for us headline here. As you can see, since January 22nd until September, October 22nd, we were in the inflationary period. There are more weights of the basket experiencing positive breaks than negatives. We had a turning point in September and since then the maximum level of pressure was actually happening around January of this year. That's a disinflation period and these lines are getting closer together, which suggests stability. But there's still no evidence in our data of a turning point For you to see a turning point here, you would need to see these two lines flip over and for a significant margin, detect more weights increasing than decreasing.

And that is not happening yet. I'm going to try to illustrate a bit more why that happened. Another way is simpler, perhaps, to look at this is to build what we call a diffusion index that not only has the positives and the negatives, but incorporates the information of sectors that have experienced no breaks. The results are very similar. The timing suggests the same thing. There have been two turning points over time. Whenever this line is above or below, you are in a period of upward or downward pressure. We are stabilizing and we are reaching new levels of strength, but there's still no evidence of a turning point there. Good. You may be wondering, what about core? Well, the story with Core is actually very similar. It's just the timing is a little bit shifted. If we look for a turning point, we did find them around November of 22. That's that's the right answer of the question I did at the beginning. But but overall, the story is roughly the same. It just happened a little bit more delayed. Great. Let's go to the another slide question, which is we are stabilizing, but are we stabilizing at levels that are higher or lower than before? And I gave you three options, actually, you can say similar than before Covid.

Uh, it's 1.5 times higher or two times higher. See how pessimistic you are. Pretty good. So you're equally split. Maybe the difference there wasn't as high enough. We're clearly on a higher trajectory, as I showed you before. Let's go back to my slides and I'll emphasize these and then compare the two ways you can look at this. What I'm showing you on the left is the turning point analysis with disaggregated data that suggests stability. Now, stability towards what you may be wondering, well, you can use the aggregate index for that, calculate these trends and then analyze the trends. So if you do that for these four periods, you can clearly see the big differences in the current trajectory, which I should say is pretty stable. It's around 3.6% that looking at our own index compares to 1.7 we had in the average between 2018 and 2019. So it's twice as high of what we were detecting before. The trend is stable, though, and I want to emphasize this will cause changes in the annual inflation rate. You can see this in this chart because of the base effects that I was describing at the beginning. We are going to see the annual inflation rate rise, even if these trends remains like it is right now. I'm illustrating that by plotting the index, as you see there on the left, I've split it so that you have the two years.

The 2020 was in blue and 2023 is in red. And then you have the months of the year on the x axis, you can see the the blue one had this slowdown, but since January of this year, the trend has been very stable. If this continues to be the case, it will have an impact on the annual inflation rate. We will see those numbers go up and many people are going to get concerned. But I want to emphasize that's not because of a change in trend right now, but rather by the base effects and what happened 12 months ago that we're going to see that continue to increase until we reach that 3.6 level. But perhaps more fundamentally, what we can do with our analysis is not just stick to the annual rate, but focus on the disaggregated trends and then see how many of the current disaggregated trends compare higher or lower relative to what we saw before Covid. So this is what this chart is meant to do. What I've done here is simply plot the share of the weights of the basket that have higher trends that before Covid, both for headline and core. If you focus on headline, the news are a little bit more pessimistic than we had seen an improvement until roughly the beginning of this year. And now things are getting back up and higher than than pre-COVID, but that most of that is coming from energy.

Once again, headline includes those energy sectors that we have seen an upward uptick in the last few months. Now, if you focus on core, this story is actually quite benign. We are still above 50%. So there are still more weights of the core basket that have higher inflation than before Covid, but we've progressively scaled that down and only 60% of those weights have higher rates. So I think that's a good news, particularly in terms of core for the analysis of US inflation. And I should say this translates quite well to many of the other countries we look at for data. So let me ask you a final question. Why was the pass through from cost to prices faster than anticipated by economists, by central banks? If you had asked central bankers, you know, how how much do you think all these cost disruptions we are seeing during the pandemic, how long are they going to take to actually affect pricing? Many of them would have said it will take a long time. We have seen that in the past. So why did we make that mistake? Why? Why was the pass through so quick? What makes this crisis different. That's another way of asking the same question. All right. Pent up demand. That's about a sudden increase in demand. Greedflation. So that's the belief that firms took advantage to some extent of this process and decided to increase their markups real time inventory.

I wonder what that means. I guess that means we are sort of monitoring data more in real time. And that's true. There's been a technological shift in many firms, not just in the way they do pricing, but how they can monitor their costs as well. But good. So you seem to think it's pent up demand. I'm going to focus on the fact that it's pent up. It conveys the idea, I want to emphasize, which is the size of the shock. This was an abnormally strong shock. And I think that's what we forgot about when we were trying to forecast the impact all these cost disruptions would have on on prices. So this is a paper we recently wrote and presented at the ECB Forum on Central Banking at Sintra. It's sort of like the Jackson Hole for the ECB. And in a couple of series of papers where we tried to do is show the importance of the magnitude of the shock to produce a pass through into prices. So if you focus on the leading models that academics use to understand pricing behaviors, we call this state dependent pricing models. They suggest and they predict they've always done predicted that a large shock will increase the frequency of adjustment for a firm that is deciding its optimal price. It's not the same to live in an environment where the shocks are small.

You get an increase in cost and you say, well, I can wait a little bit longer, but when the shock is big plus it has been accumulated for so long, you have to start making pricing decisions more frequently and that frequency translates into higher pass through. What you're looking at here is the actual frequency numbers we can get from our series in the food and beverages sector. In the past two years, there was a dramatic increase increase that actually coincided with the war in Ukraine. That seemed to be a point where many firms perceived that the changes in costs had been large enough that they needed to change the way they were doing their pricing. And that has continued to be quite high ever, ever since. This leads to higher pass through. And you know, the if you read the paper, you can see the modeling side, how quick this pass through can actually be given differences in shocks. We're assuming here a 20% increase in an energy shock. And there are two lines here. The red line is what we call a time dependent model. This is a sort of kalibo or time dependent models that central bankers have used for the last 15 years. There are models that work really well in stable environments. There are models that tend to assume firms adjust their prices once or twice a month at a given interval, and that's perfectly fine for normal times.

But the truth is, when the shock is large, you have to understand that the model is completely wrong. It will give you the wrong prediction about the degree of M.A. Now this is more of an academic discussion. I know and focus on the past. The question becomes what can it tell us about the future? So the theory we still have to take this in the data. We're going to see how this plays out. But the theory suggests this pass through is not and this behavior is not symmetric. So there's less incentive to pass through cost decreases. That cost increases. And the reasons for that, the micro reasons are that for a firm, having high price is actually better than having a low price. Sure, if your price is above what it should be optimally, you're going to be losing some sales. But for the items that you sell, you're actually making a good profit. Whereas if you're below, you're actually selling goods below your your actual cost that you're going to be losing money. And then there's another more dynamic question. If there is inflation, positive inflation and you know your price, your optimal price will eventually be higher than what it is today, There may be no need to actually drop your price today. You can wait until inflation takes your optimal to the level you're at today. So all these kinds of behaviors predict according to the theory that we should not expect for the same reasons we saw a quick pass through on the way up.

We should not expect the same pass through on the way down. So that explains a bit some of the more recent dynamics and reasons, at least theoretically, of why this makes sense. So I'm going to emphasize some key takeaways. First, it's important to remember that retail stockouts and suggest these improvements in supply have helped achieve disinflation without recession. But we're already at pre-pandemic levels, so we have to ask what is going to happen moving forward? Second, the structural break analysis I showed you today suggests we have broken inflation in nearly all the countries we look at in the US. It happened relatively early in 2022, but we are stabilizing for now at rates that are higher than pre-COVID levels. Core, though, looks better than most people. Argue and I want to emphasize that and we have different stories across countries. China, we had an early slowdown in Europe. We have a very late break and a lot of divergence that is worth analyzing. The size of the shock suggests that explains a lot why we had such a quick pass on the way up, but suggests the path on the way down has to be slower and thinking of next shocks and what they will produce to the to the dynamics globally on the next few years. I want to emphasize I have good news first, which is that we are launching a China aggregate index.

We have always produced a food index and a supermarket index. We never had the ability to cover as much Chinese data online as we wanted to produce an aggregate index. But we have been doing a lot of effort in the last few years and we are going to be launching and replacing our traditional food and supermarket indices with the index that you're looking at here in just a few months. This is a consistent or comparable to the aggregate CPI of China. And you're looking here at the index, the monthly rates in the annual rates. I'm just since we're talking about trends, I just want you to focus for a second on what you see there in the index of China since the beginning of 2022. There's a lot of ups and downs if you look at monthly and annual rates, But the trend of inflation there is quite deflationary. And that is a problem not just for China but for the rest of the world. It suggests that there is a lot of downward pressure on prices and problems with the economy in China and that will have an impact on globally. We will be monitoring all this information in real time with our price indices and telling you how we think this is going to affect dynamics moving forward. But let me stop for a second and then open it up for questions. Thank you very much.

Speaker2: Thank you, Alberta. So there are a couple of different ways that you can you can ask questions. So on the Slido that you've great questions that you've already been using, just jump the tab across to Q&A rather than polls and you can submit them and they're going to appear here some Wow, someone's a quick typer. That's great. Or actually, goodness me, with everyone in the room, you can also raise your hand. I just ask that you that you wait for the mic. So why don't we actually let's, let's take let's take a live one since we're all here. There's a lady here I think with her. Hand up. Dave, you got.

Speaker3: Hi. Thank you for the presentation. Quick question. Where do you think inflation will land? Like? Do you think it's going to be at 2% for the US or do you think we are at a structurally little higher inflation? Thank you.

Alberto Cavallo: Great question. Right. So there's there's part of me that says let's focus on what the data shows today. I can tell you we're still very stable at that 3.6% rate. So in the near short run, I think that's going to be the rate on the headline. On core, we don't produce a core index per se, but I showed you 60% of the weights of the basket are at a level that is higher than before. If I take an average of that and compare, it actually looks better than headline because we have an average level of 3.2% on an annual basis. If I extrapolate those trends compared to what we had pre-COVID in that subset of sectors that we can monitor online, that was around 2.4%. So it is higher than before, but that's a very short run. And under the assumption, by the way, that we keep on on this on these trends, the data tells me they look quite stable. So I have no reason to believe they're necessarily going to change in the very short run. Now, of course, the question becomes what happens in six months, a year from now? That's what you're interested in. And the truth is that will depend on so many things we have. The way I think about it is what has been a change that is structural enough to explain a consistently higher rate of inflation moving forward. Because you can think of changes that for a while will produce high inflation numbers. There are disruptions in shelter, for example, that the pandemic brought about. Those may last a little bit longer. But can they sustain higher levels of inflation moving forward? I'm not so sure about that. For example, what I am sure is that, for example, supply disruptions on the on the good side that appear to be more longer lasting than before, those have gone away now. I've seen many things brought about by Covid eventually returning back to normal. So I'm overall optimistic that we'll get there finally, I should say. No, I want to end on a positive note, so it's okay.

Speaker4: All we're going.

Speaker2: To remember is the 3.2% forecast. We've got that on video, so we're good. So got some great questions coming in. I'm going to try and group a few of them together just in the interest of time. How many months does it take to constitute a trend in your analysis? What length of trend do you recommend using to gauge a lasting shift?

Alberto Cavallo: Well, that's a great question. So there's a parameterization of the algorithm. We do it based on the behavior of the series that we're focused on or the country we're focused on in the US. We can detect a break if it's strong enough within just a couple of weeks. But that's a choice also of the person running the algorithms. If you're really looking for short, quick trends, you can you can do that. I've tried to focus on trends. For example, we trim the data by about a month, so we cannot detect anything that has happened in this month, which I think is reasonable. You shouldn't overreact to what you see in a single month. But in the US, if the trend changes is big, we can within two months we can we can be confident about it. Now it's very different if you ask me about other countries. So for example, in Argentina, where I'm from, we just had a dramatic change in the trend and we can detect that within a couple of days because there was a massive devaluation and then prices start rising quite dramatically.

Speaker2: Brilliant. Okay. So a question about the word cloud. Interestingly here, so Greedflation was the second largest term in the word cloud. You identified companies keeping prices high in your micro analysis. Can you expand on that?

Alberto Cavallo: So that's a very good question. I'm actually writing a paper with a very large manufacturer, global manufacturer, to try to understand what happened with their markups over time. There's this idea that markups went up a lot and that helped them and explains to some degree inflation. What we're finding is that firms did experience variations in markups. Sometimes the markups went up, but not for reasons you might think. At the beginning of the pandemic, some markups went up because costs were falling, energy was falling dramatically, and firms were not changing their prices. So they're actually their markups were quite good in the in part of 2020, particularly after demand started to to recover. And they sold a lot of goods. I think like most economists, this is not a driver of inflation, but certainly understanding the incentives of firms in when to do pricing is key to understanding why these dynamics play out in certain ways. And that's what I'm trying to do with the work that I do, For example, on pass through, understand why is it that firms took a while to react, why they reacted so much in the past year, and what I think will happen moving forward. So I'm not particularly in the camp that think Greedflation is what has been driving this. But certainly the attempt to recover margins will explain some of the persistence in these inflation rates that we see moving forward. And many of them may be catching up. Others, of course, some particular sectors may have experienced a big increase in demand and no change in cost. So they have actually received quite a boom in this year. But overall, my sensation is it's not the reason why we saw inflation. Inflation has to be attributed more to policy and the effects on demand and to supply disruptions. Clearly putting upward pressure on prices, as I've shown you. And then understanding the behavior of firms is certainly important for the dynamics moving forward, but not a major driver.

Speaker2: Okay. So we've got a couple of minutes left. I'm happy to take a Julia.

Speaker5: Do. Do you see any patterns in the distribution of price changes across sectors in the more disaggregated data that you showed? In the sense that in the pre-COVID environment you might have had much more similar changes, whereas now there have been big demand shifts across sectors and you may be seeing a greater distribution of change in relative prices. Have you observed anything like that?

Alberto Cavallo: So we in this work I did recently for the ECB and the pricing behaviours we detected as a mild increase in the standard deviation of price changes. So there was certainly some volatility more than before, but it's not the major driver of, of or the main pricing behaviour that has changed. It's more become a story of how frequently firms make these decisions rather than how dispersed these price changes end up being. You can, as inflation is rising quickly, expect some relative prices to be out of whack and not particularly well aligned anymore. But there's nothing that jumps out as being very dramatic at this stage.

Speaker2: Oh, yeah. One more. The back.

Speaker6: Thanks, Ali. Yeah, I was hoping to get your reaction to two statements. One is every time year over year, CPI has been above 5%. It's led to recession. That's the first statement. And the second statement is the quality adjustments that have been made to CPI that started in the 1990s. Are those causing more problems today due to the fact that the pace of innovation is that much more rapid?

Alberto Cavallo: All right. So the first question I think you're absolutely right. But we we should be always careful of looking back and assuming the same thing we have seen in the past will happen today. I think a good way to frame that is to think what is different today. And certainly, as I've shown you at the beginning, Covid was a different crisis because of the impact of supply. So to the extent that supply can recover, we may not need to see a recession and we can bring it down from above 5% compared to the last few crises. Major crisis in the US and in other countries that have been more demand only. And sorry, the second question I forgot.

Speaker4: Yeah.

Speaker6: Adjustments to CPI and whether or not that's causing more problems due to the.

Alberto Cavallo: Right? Yeah. So I don't think it's necessarily affecting the quality of the indicators, No. And lately it can have an impact more on the long run. You know, if we're talking about five, ten years, it can lead you to a price level that is perhaps five, 10% in some categories different. And you would be missing that. But in the very short run, I wouldn't think it's something we need to necessarily worry so much about. I should mention, in fact, if you compare our data to the CPI, the way we do quality adjustment is different from the BLS. The BLS goes and tries to measure how much of the price change that we see is actually related to quality or not. They do those hedonic adjustments. We rely on the fact that we have a lot more data and then kind of calculate an average price change for a very large basket, which gets us close to or approximates that estimate that they're doing. And they're essentially very different techniques. They yield relatively similar results. So I don't think for short term dynamics, it's something we should be worried about.

Speaker2: So I'm sorry to say we've got about 5 or 6 terrific questions on Slido, but we are out of time. But what I will tell you is that we will get to those questions. We'll find a way to get the answers to you. But, Alberto, thank you so much.

Alberto Cavallo: Thank you so much.