Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=

# Trade Fragmentation, Inflationary Pressures and Monetary Policy

Authors:   
Discussant: None  
Video: https://www.youtube.com/watch?v=IghWOJTC51A&t=0s


## Talk (00:00:00 – 00:25:00)

[00:00:04] Hi everyone. Can you hear me? Assume. Yes. Okay. Well, uh thank you very much uh to the organizers for putting our paper on the program. Uh this is joint work as you can see with uh Ludvika Ambrosino who's a PhD student at LBS and  
[00:00:18] Sana Tenedo at the LSSE. Um how do I advance the slides?  
[00:00:25] Oh yes. Okay. Thanks. Uh thank you. Uh so let's start with some context for our paper. Um as you can you know as you can tell in recent decades global trends  
[00:00:37] have shifted uh quite noticeably. So um as you can see so yeah this long post-war increase in trade openness has stalled as you can see from this figure which shows uh global trade to GDP plateauing after the global financial  
[00:00:52] crisis. Moreover, we also know that's underlying many of this uh many of these trends is that trade is increasingly being influenced by geopolitics here.  
[00:01:02] And so we have a fragmentation index uh that's put together by Jesus Fernandez Verde and his co-authors. And it shows that since the global financial crisis, fragmentation has been increasing gradually, spiking up around the  
[00:01:14] pandemic and the war in Ukraine. Okay. And so these trends are also likely to persist. There are many um nor many things that are being normalized front shoring and realignment of trade towards geopolitically friendly countries. Now the key takeaway that I want you to take  
[00:01:29] from these first two figures is that this is not necessarily going to be reflected in a fall in trade openness but a realignment of trade to uh from the most efficient suppliers to the more geopolitically friendly ones.  
[00:01:44] And uh so for many countries then trade fragmentation driven by geopolitics is likely to lead to higher imported goods prices and lower real incomes. And against this backdrop um two concrete questions that we will address are going  
[00:01:58] to be will fragmentation lead to a high inflation environment and what would be the monetary policy response that's needed to keep inflation at target. Let me give you a preview of our answers.  
[00:02:09] Okay. On the first one, will fragmentation lead to a high inflation environment? So in principle, no. This does not mean well in principle this does not mean that central banks need to change their remits. But a better way to rephrase this question, especially given  
[00:02:22] the policy panel yesterday, is um will fragmentation lead to higher inflationary pressures even if inflation ultimately returns to target because of actions taken by the central bank? And the question to this um the answer to  
[00:02:36] this question is it depends. Okay, so we show that in some scenarios, for example, when fragmentation is very front-loaded. So you can think something similar to what happened with the war in Ukraine, this can lead to a stagflationary scenario for policy  
[00:02:51] makers. So with a spike in inflation and a fall in outputs would leading uh lead to a short-term trade-off for policy makers. But this is one possibility.  
[00:03:00] While this is one possibility, another could be that if fragmentation is gradual and anticipated, so kind of like the unwinding of uh globalization, this can lead to stagnation. So with  
[00:03:13] lower real incomes and lower well domestic disinflationary pressures and this leads to the second question, what is the monetary policy response then needed to keep inflation at target or how will equilibrium rstar respond? And  
[00:03:26] here of course the question it follow well the answer it follows that it also depends it really depends on how demand responds to permanently lower incomes and if fragmentation leads to a sharp fall in demand this could in this could  
[00:03:40] lower our star. All right. So, um just very quickly over the literature, we build on some classic um models in the literature. Um Benino  
[00:03:52] and Ben go andbe Galian and Monachelli uh Santra and Deali also some work by uh John Carlo. Um I want to highlight of course this builds on the earlier literature on globalization  
[00:04:05] um and the newer literature on uh the impact of tariffs which Alberto has contributed to and uh on the empirical front I just want to highlight one strand. Uh so many people seem to  
[00:04:16] attribute uh globalization to um a fall in inflation well sorry global integration to a fall in uh inflation and um on this front uh there's some interesting empirical work by authors at  
[00:04:31] the ECB so atonasian balati uh has contributed to this with his co-authors at the bank of France also at the bank of England we have Lewis and Salah and they show that they make a compelling case that it was really monetary policy and the introduction of targets that led  
[00:04:44] to a fall in inflation. So this link a causal link is elusive. Uh what we're not trying to challenge is that of course it led to a fall in tradeables inflation but what happens on the demand side is also quite important as well. If monetary policy keeps the mon nominal  
[00:04:59] anchor um it's not necessarily the case that um fragmentation leads to higher inflationary pressures. Okay. Um okay just I'm going to go over the model uh a  
[00:05:11] bit. Um might be a bit repetitive. I'm going to give you a summary and then a a bird's eye view of the production structure of this economy and then go into detail. But I've heard that the model is not uh small. So maybe it's  
[00:05:25] better to air on the side of being a bit repetitive. Uh just to summarize uh we have a small open economy model uh newian model with two types of households unconstrained and constrained households. Um the unconstrained are  
[00:05:38] your usual representative agent households. Okay. So they maximize utility over consumption, labor supply, asset holdings uh subject to their budget constraints. We also have constrained households who spends all of their disposable income within a period.  
[00:05:52] Um they supply labor. So this is labor income. Firms are standard. They maximize profits given production technology. We have two types of firms.  
[00:06:00] Well, two sectors. We have a non-tradable sector where there's monopolistic competition and sticky prices. And uh we have a tradable sector which yeah domestic tradable sector um  
[00:06:13] where prices are set internationally um other prices are flexible.  
[00:06:18] Okay. Um in so in this home economy which is the one that we analyze we take the rest of the world's dynamics to be exogenous. This home economy trades consumption goods and imports foreign goods. Um they trade also domestic and international bonds and it is the  
[00:06:33] unconstrained households that undertake this. Uh constrained households again they have no access to financial markets. Um the unconstrained households they trade these riskless domestic and foreign bonds and there's a quadratic cost in changing the real asset position. Uh we have a monetary policy  
[00:06:48] rule where uh they respond to deviations of CPI from target.  
[00:06:53] And this is just a bird's eye view of the economy. what I've discussed it's uh the picture of the small open economy that we analyze two types of households they supply labor to domestic production and they consume domestically there are  
[00:07:07] two sectors in the economy tradable sector exports to the rest of the world um unconstrained households borrow they can borrow from the rest of the world and the import well the import price shock that we're going to look at is  
[00:07:20] going to affect uh the domestic economy through the left side of the chart which shows um an effect on imported inputs in non-tradables production. You can think of this as energy and it also affects  
[00:07:32] the final consump uh final consumption good of part of the consumption basket.  
[00:07:37] So the consumption basket of households will have non-tradables, tradables and then tradables will be home tradables and foreign produced tradables.  
[00:07:47] Okay. So let me just go into a bit more detail about the various uh blocks in the economy. Then these are household preferences. They uh maximize expected lifetime utility. Uh consumption basket is CS aggregate of tradable and  
[00:08:01] non-tradable goods. One minus v sigma is the share of tradable goods and consumption and in turn uh tradable goods is composed of home tradable goods and foreign tradable goods um one minus theta is the home bias of the economy.  
[00:08:16] So a more open economy is going to be more adversely affected by these um uh foreign shocks.  
[00:08:22] uh non-tradable goods are uh a CS aggregate of individual varieties and last epsilon is the last of substitution between these varieties.  
[00:08:33] Now the prices just follow from the nesting that we've introduced aggregate price level composed of tradables and non-tradables prices. Uh tradables prices composed of home tradeable prices and foreign tradable prices and  
[00:08:46] non-tradable goods prices is a composite of the individual varieties. PFT is what we're going to shock in the first two simulations of this uh um yeah so I'll take you through a few simulations to  
[00:08:59] show um the impact um for households we have uh the unconstrained ones these are your standard uh representative agent ones oh sorry these are the  
[00:09:13] yes these are the unconstrained households they have access to uh foreign and domestic bonds budget constraints and v real variable will show um that they can borrow uh domestically and from abroad. There's a cost in changing the portfolio position.  
[00:09:28] They receive firm profits. They have labor. They supply labor. So they receive labor income and this is what they consume out of. Um so for the unconstrained households, of course, they have optimal labor. Sorry.  
[00:09:41] Uh so this is still the unconstrained households. Therefore, sort of conditions give you optimal labor supply, oiler equation and the UIP condition which holds up to the cost of changing these uh bond positions. Uh for the constrainted households, they have a  
[00:09:55] much simpler budget constraints. Uh they supply labor optimally, but they consume all of their labor income within the period.  
[00:10:02] Um now I'll take you through the firms. Uh starting with the non-tradable sector. So this is the domestic sector.  
[00:10:09] Um production is called Douglas. Uh so it's it takes intermediate inputs um and labor domestic labor and they take uh the price of these inputs as given. So  
[00:10:22] uh with labor with wage wage rate W and imported input with the foreign price uh there's monopolistic competition in this market and there's sticky pricing roenberg uh pricing and so there's uh  
[00:10:35] adjust well yeah uh production is there's going to be some efficiency loss there because of the price of uh of adjustment.  
[00:10:44] Uh in the tradable sector firms uh produce just using labor and um the sector is internationally competitive takes global prices as given. Labor is used in both sectors  
[00:10:58] and the calibration is uh there's some more detail in the paper. I just want to highlight that we deviate from log utility and we assume unitary elasticities of substitution. Um we think this is a reasonable assumption for globalization which takes place over  
[00:11:11] a period of time but uh we also have varying elasticities of substitution in an extension and I'll tell you a bit more about the the results from there.  
[00:11:21] Uh so just to show and what we want to highlight with this is that um we want to highlight with all three different types of fragmentation scenarios that the supply side is going to be constrained uh with this uh trade fragmentation. But  
[00:11:35] what really matters and what we want to highlight with the different scenarios is that the demand side impact also matters and it's going to differ in each of the different scenarios. Okay. So for example the first one that we have is the gradual scenario and this is the one that leads to stagnation. This is one  
[00:11:50] where the price of imported goods increases gradually and permanently stabilizing at a higher level in the medium to long term. Front-loaded fragmentation on the other other hand takes place immediately. So the price of these goods just increases permanently  
[00:12:05] and uh immediately. And lastly, there's also a fall in tradable sector productivity. So TF in the in tradable sector falls persistently.  
[00:12:16] And so I'll take you through first uh the gradual scenario and I'll do this in a rank model because it's cleaner and it's easier to see the intuition here.  
[00:12:24] So we have a gradual um gradual increase in the foreign price level which you can see from the figure on the bottom and this increase is anticipated.  
[00:12:34] Consumption falls be in anticipation of lower real income. So these are all representative agents. They adjust consumption in response to permanently lower incomes. Labor demand falls while labor supply increases. This wealth  
[00:12:47] effect and real wages fall. So there's less consumption and more labor effort given the worst terms of trade in this scenario. But I can show you what happens to prices then.  
[00:12:57] So the fall in demand is reflected in the fall in the natural rate of interest here. Inflation which you should look at is aggate CPI inflation that falls because the domestic components of inflation fall. So domestic components  
[00:13:10] of inflation are given by non-tradables inflation. Can see that it falls uh over a long period of time and ultimately recovers. Um but this more importantly you have to contrast it to imported inflation. So there's a spike up in  
[00:13:25] imported inflation and this foreign price level increases but this is not enough to outweigh the fall in non-tradables inflation. So this is a situation of stagnation. Monetary policy actually has to loosen in order to bring inflation CPI inflation back to target.  
[00:13:40] Now contrast this to the case where um there's a front-loaded increase in import prices. So foreign price level you can see on the the last panel increases immediately and permanently  
[00:13:52] and consumption and real wages fall but the natural rate of interest does not change the so CPI inflation increases so aggate CPI inflation is what we should be looking at this is even though the domestic components of inflation fall so  
[00:14:06] we also have a fall in non-tradables inflation here but it's really shortlived it recovers immediately almost immediately and more importantly it's not enough to outweigh the high spike in imported inflation. So this is a scenario that results in a trade-off  
[00:14:21] for policy makers. It's stagflationary. Aggregate CPI inflation increases, activity falls. And this is one where monetary policy ends up needing to tighten in response to mon to the spike in CPI inflation.  
[00:14:35] Now another scenario that could happen is that uh tradeables TFP just deteriorates and stays permanently lower. So we have here a a fall a gradual fall in domestic tradables TFP sector. So this is a purely domestic  
[00:14:49] shock. Consumption here falls as well. Natural rate of interest also falls.  
[00:14:54] There's an initial increase in CPI inflation. Uh but it also it follows by a permanent fall. Monetary policy needs to tighten first and then loosen.  
[00:15:06] So uh next I'll take you through the scenario where we introduce some hand-to-mouth households in a tank.  
[00:15:11] Okay. So uh the initial motivation for this scenario was uh to introduce some hand-to-mouth households which cannot smooth um they cannot react to the import price shock in anticipation.  
[00:15:24] Okay. But it turns out that there are also so this lessens the anticipation effect but it turns out there are also demand spillovers as the unconstrained households pull back in consumption.  
[00:15:33] This affects labor income of the unconstra of the constrained households.  
[00:15:37] So there's not you can see the difference uh the fall there's a slightly larger fall in consumption. Uh you can see that the constrained households decrease consumption by more uh they're unable to smooth but um these  
[00:15:50] two effects result in pretty similar outcomes for the prices. There's a larger degree of reallocation from the non-tradable sector to the tradable sector here because of the constrained households but other than that we get  
[00:16:02] similar um outcomes for inflation. So our results are robust to the introduction of these households um similar uh similar balances as in the rank and this is still a case in which um the gradual scenario is one in which  
[00:16:17] it's stagnant. So you have a fall in inflation, a fall in activity and a fall in the natural real rates. Okay. uh similar for a front-loaded scenario.  
[00:16:26] There's slightly larger fall in consumption and reallocation towards tradables, but same uh leads to fairly similar outcomes across tank and rank and not much difference in inflation or the natural real rates. This is still one that leads to stagflation, the  
[00:16:41] front-loaded scenario and same uh for the tank uh TFP shock. Uh we don't have much difference um between the rank and the tank.  
[00:16:50] Uh the degree of openness is a bit more interesting. uh we vary the degree to which this uh domestic economy is exposed to these uh price shocks, these foreign price shocks. And uh so just the  
[00:17:03] major takeaway from this is that yes, there's higher exposure as trade openness increases. There's more adverse outcomes in the more open economy which is the blue line and blue line and then followed by the blue uh the black dashed  
[00:17:17] and then the red economy is the most closed one. So the so larger falls in consumption wages in the more open economy and more reallocation towards home tradables. Uh interesting uh baseline is the nearly closed economy  
[00:17:31] the red dashed line. So there imported inputs mainly affect um non-tradables production and then the additional um the the blue and the black dashed are what happens when we increase the consumption basket to have more foreign  
[00:17:44] tradables. So the rest what's coming from that is the demand side effect from the effect on the consumption basket.  
[00:17:51] Okay. So otherwise in the uh red economy it just affects the price of imported inputs for non-tradables production. You can see there's a fall non-tradables output and employment um and uh yeah so  
[00:18:04] a fall in the real wage as well. Okay so similar outcomes for yeah so this is still the gradual case. Higher exposure leads to larger domestic adjustment in these prices leads to larger fall in domestic components of inflation and  
[00:18:18] natural rate of interest. So this is still a a stagnation scenario but it's more st more so in the case of an open economy. Now in the case in which we have a front-loaded increase in import prices leads to a more difficult  
[00:18:31] trade-off for policy makers. So you have more adverse outcomes and lower demand lower real wages but higher CPI inflation. But importantly natural real rate does not change here as well.  
[00:18:43] Okay. And for the gradual TFP shock this is a purely domestic shock. Okay. So uh openness matters a bit less. Uh we have a fall. So it's not shown here but in the paper you see there's a fall in output. Uh consumption is actually uh  
[00:18:56] the intuition is a bit more clean in the rank. Consumption here actually is more adversely affected in the open economy because of um the distribution of profits. profits are actually more adversely affected. Uh so the  
[00:19:11] unconstrained households um take a bigger hit uh in the more open economy.  
[00:19:16] Um but this is a case where we wanted to show that uh for purely domestic shock actually being open helps. So it actually the outcomes on outputs are less adverse because you can diversify away from um the affected home tradable  
[00:19:30] sector. Okay. So we have a few extensions in the paper. um we just show that uh for example what would you think of um so what happens when we have more uh constraints on domestic supply so for  
[00:19:45] example if you're more dependent on foreign inputs for production when there's wage stickiness um and so in these across all of these scenarios um all these extensions also with more flexible prices and non-unary less  
[00:19:59] issues of substitution we get the same conclusions across our scenarios gradual is still stagnant um front-loaded is still leads to a trade-off and TFP leads to a mild uh increase in inflation. Um but the demand  
[00:20:13] side effects are preserved uh across the scenarios and I'll just give you some quick intuition as to why that is uh why we still have our conclusions holding.  
[00:20:21] So uh when there's a higher share of foreign inputs in production um this leads to a larger fall in consumption real wages because of in response to increase in foreign prices. So the foreign the increase in foreign prices is a a tighter constraint on domestic  
[00:20:35] supplies. Import prices increase. There's some factor substitution towards labor to the extent that that can happen with cobb douglas but it's not enough to stimulate aggregate demand. And moreover um so we get that's the gradual scenario is the demand side effects are preserved  
[00:20:50] there but um it seems to matter a bit more for the aggregate uh TFP scenario where it's a little bit less inflationary because now that uh non-tradable's production depends more on inputs and less on labor. It's less  
[00:21:03] able to absorb um absorb labor uh from the effect of import prices on home tradables. Okay. So this leads to less inflationary pressures in the non-tradable sector which leads to less  
[00:21:16] aggate uh CPI inflation in general. Wage stickiness is a bit interesting as well.  
[00:21:22] Overall it moderates the fall in wage inflation but it also leads to a larger decline in employment. So quantities adjust and so labor incomes are preserved in the scenario. Uh the gradual scenario is less disinflationary because for a given fall in outputs  
[00:21:36] aggregate CPI is actually the same. Um so a output falls by more but CPI is the same. So this is a more um a less disinflationary scenario. It leads to worse trade-offs in the front-loaded scenario. Uh so aggate CPI is actually  
[00:21:49] higher in impact and more persistent and leads to worsens the policy trade-off. U more flexible prices. One of the frictions that we have in this model is that oh thanks is that a non-tradable sector is the sticky sector and um the  
[00:22:03] inability to adjust prices leads to adjustments in quantities. Um so actually if we have more flexible prices this increases disinflationary pressures in the gradual scenario and it actually lessens the policy trade-off in the front-loaded one. We also have one with  
[00:22:17] different elasticities of substitution leads to less stagnation the gradual scenario but overall um our conclusions across the three different scenarios are the same just uh the underlying mechanisms are are slightly different.  
[00:22:29] Um so just to quickly conclude uh fragmentation may lead to higher import prices and lower supply lowering real incomes. Now the impact on domestic aggregate and CPI inflation really depends on how demand adjusts to lower incomes which in turn depends on how  
[00:22:44] fragmentation materializes. So we show that in the case where it's gradual, it could lead to stagnation with low real incomes and low inflationary pressures and this is where monetary policy might actually need to loosen. Front-loaded scenario leads to tradeoffs uh or you  
[00:22:58] stagflation. Persistent falls in tradable TFP can lead to stagnation and lower interest rates. We show that depends on the calibration um can also lead to a moderate increase in inflation. and how monetary policy response uh should respond really  
[00:23:12] depends on the balance of supply and demand and the policy direction is a pretty ambiguous. So some next steps we can also consider other types of fragmentation scenarios. For example, this is anticipated could be that you  
[00:23:26] have unanticipated sustained increases. Um we abstract from lags in policy transmission and the fiscal policy response. Of course you can also have um non-rational inflation expectations.  
[00:23:38] Um so just quickly and briefly just to step outside the model and think of like bigger picture implications.  
[00:23:46] Um I think it's it's just pretty obvious that you know to tackle real side geopolitical issues requires perhaps real side policies um and um monetary policy might need to you know in the in the events where these  
[00:24:00] are absent monetary policy might need to um address the residual effects of this.  
[00:24:06] Um but some real side policies to prevent, mitigate or cope with these impacts of u of geopolitical developments would be investments in diversification of technology focused on uh low substitutability inputs or  
[00:24:20] technologies. Um deeper integration with low-risk countries to reduce exposure to these shocks and inventory bases to prepare for shortages in critical inputs. And with that that's pretty much  
[00:24:33] it. Thank you very much. Thank you.  
[00:24:39] >> Looking forward to John Carlos to discuss.  
[00:24:41] >> Thank you, Johnny. Lots of material to think about. I count on Jan Carlo for making me at least understand some of the channels better.  
[00:24:59] Thank you for inviting me. It's nice to be back in this room. I've been here for many many years in the times the monetary institute. So um  

## Discussion (00:25:00 – 00:38:42)

[00:25:10] how does it work? Oh yeah.  
[00:25:16] So uh I I like this paper. I mean it shows basically how you can think of good policy drawing on good economics.  
[00:25:24] So I'm going to do exactly what K asked going back to very simple principle. So the question is clear. How will track fermentation impact on monetary policy?  
[00:25:34] We understand important intermediate prices going up is a production cost in inflationary pressure. Uh there is also a problem with employment.  
[00:25:43] But fragmentation also impact wealth and demand. So in principle we could have a scenario in which wealth effect and demand prevail. So low natural rate, low inflation or a scenario in which supply  
[00:25:57] effects prevail. In that case, we have a stackflationary scenario.  
[00:26:01] So you could do big models or you could do a pretty stylized although sophisticated model and use economics 101 which is terms of trade versus productivity shock current versus  
[00:26:16] anticipated. So why is that interesting? Well, you know, in a way um we can think of intermediate inputs uh prices coming from disruption which is terms of trade  
[00:26:28] shock and you can think of problems linked to technology, innovation, migration that may bring in an effect on TFP. So that's the motivation.  
[00:26:41] So uh what are the theoretical foundations that uh very basic that shape the analysis? One is obvious.  
[00:26:48] Sorry to say this. Shocks are shocks. They have demand and supply effects.  
[00:26:53] There are no supply shocks and demand shock. There are shocks fundamental with demand and supply effects. Within complete markets, shocks may have significant wealth effects driving demand. And they drive demand in two  
[00:27:05] ways. The first one, if I count only my uh uh real wages, for example, uh ready price  
[00:27:16] changes affect my wealth and and demand. Income effect come from large changes in relative price that can be endogenous or exogenous and of course reflect anticipation or  
[00:27:31] real income in the future. No. Uh notice the difference. uh usually when we think of real relative prices we think of shortrun low elasticity that may come into creating issue with  
[00:27:46] JVC disruption the elasticity is low all of a sudden I don't have an intermediate input prices move a lot that feed this kind of supply price factorial incomes when instead we think about the long run  
[00:27:59] anticipated shocks we think that we we'll move out from GVC as we know into something else so This is may be higher but this makes any kind of output effect in the future even more important for me because the price does not change much the output may change quite a bit for  
[00:28:14] example with TFP and the last thing if you have multiple sector and multiple agent which is if you really face reality there is something more in the policy tradeoff than just aggregate  
[00:28:28] output aggregate employment because there is relative price misalignment misalignment in sectors misalignment in demand or demand funding balances and this all comes into your lots function.  
[00:28:39] So Luca is here, Simon Lloyd is in England. We have been working a lot on this wealth effects uh with a lot of papers including the new macroeconomics book by Perus atal chapter 23. Please  
[00:28:52] read it because it's all about you cannot do really international monetary policy meaning you cannot do monetary policy if you don't understand this kind of international aspect and dynamics of  
[00:29:04] u u wealth fact that I rediscovery even in close economy when we talk about canian supply shock is the same thing and also about the relevance of this trade elasticity debate for macro comes  
[00:29:18] from this work from in which you know that they use this very low elastic it is to show the actually ready price can impact when you have ant-to-mouth consumer quite a bit and last bit the  
[00:29:32] loss function you know you need to track sector reallocation really price misalignment when you do monetary policy because those are relevant to people they're relevant they're costly to reallocate h agents or people across  
[00:29:46] sectors so let me tell you very briefly the building block of the model there is economic structure with UK and Roterberg two sectors one is competitive flex price the other one is sticky price the  
[00:29:59] sticky price is non tradable and it only uses intermediate uh inputs you have lots of ballast so the Fed actually were crucial in explaining the the pound depreciation after Brexit the other  
[00:30:13] models could not really fit that thanks to Silbana there is a small open economy so you can take ready price as given and there are two two types of agent I explained by the way tanker rank in open economy makes no sense because you have  
[00:30:27] more than one representative agent. So you could have a national representative. So you could call it n tank or you could call it tank which is two rep representative agent. We ask judg the tank idea or tank. So this is the  
[00:30:44] tank and the titank model. So let um I'm I'm some problem I I you know like in the way that the paper  
[00:30:55] um uh it could be at some point explore more about elasticities you could put more intermediates in the model more nominal rigidities like the previous paper and um a little bit discussion on  
[00:31:09] natural allocation but it's fine I'll talk later this remember there is one sector with one nom rigidity and Secondary is incomplete market. So let me tell you the story the answer to  
[00:31:20] Kiara. So suppose I have a permanent adverse terms of trade effect occurring at once in a flex price equilibrium you will have a negative wealth effect driving consumption down labor supply up  
[00:31:34] and real wage down. It's all like a general equilibrium gen equilibrium responses. What is important is that there will be a change in the relative price right which will drive reallocation in demand and employment from the non-tradable to the tradeable  
[00:31:48] sector because become less efficient and more costly. Now with sticky price actually you could act you could just target this thing how do you do that because we have only one rigidity PPI works if you define PPI as targeting the  
[00:32:03] core inflation with the sector with nominal rigidity. So what you do you're basically you want to prevent sticky price firms from changing prices because that is a waste waste of money no into  
[00:32:15] the uh costly adjustment waste of resourcing to the cost adjustment of prices. So you let the CPI accommodate ready price. Okay you call it CPI inflation but it's actually an official  
[00:32:28] price movement. So the lesson is that if you do CPI target in this world is not good, right?  
[00:32:36] It's too contractionary. You forget that there is a a an efficiency price adjustment. You try to lean against it which means that basically you contract consumption too much. You don't let the price to adjust and you also get  
[00:32:51] inefficient excessive real allocation of employment in the short run. So if you look at this figure, if you look to the right of H graph, you'll see the long run. If you look to to the left at the  
[00:33:04] beginning, you will see excessive excessive reallocation of the excessive reallocation across sector where you overshoot the reallocation.  
[00:33:17] You get more too much reallocation to the tradeable than then comes back to the long run. And remember relocation is costless in our model. Not in reality.  
[00:33:28] Not in reality. And my answer is that it may be the same order of magnitude if not more than than inflation. So you you want to you cannot ignore relative price. You cannot ignore this kind of dynamics as policy maker. By the way, it  
[00:33:42] would be nice to to plot the relative price here in in the this is a complaint I have. The only complaint with Jenny just plot the dynamic of the relative price here. Uh so now suppose again in  
[00:33:54] tank that the the um the problem is coming it's gradual not tomorrow we'll have this adverse terms of trade shock well you know immediately you have a negative wealth effect you understand  
[00:34:08] that you're poorer tomorrow so that means dropping consumption a little bit of adjustment labor supply deflationary pressure okay so if monetary policy fail to understand that fail to expand enough  
[00:34:23] to keep consumption up today. That means the firms that get costly adjustment price will start dropping prices wasting resources. You don't want that. Okay? So you want to avoid the misalignment  
[00:34:36] related price. You want to avoid the excessive allocation. You need to push a little bit up. Okay? And and the reason why is good is because in a way tomorrow is the problem. Today is nothing. In this model there is no capital  
[00:34:50] anticipation. So basically you wait until tomorrow to adjust to today you just want to keep things as they are.  
[00:34:58] Now again a subotimal monetary policy less deflation to face in up front and that means that you are stuck with a period of expansionary policy just to bring inflation back to target right you don't take advantage of the flexibility  
[00:35:13] of PF and PN in the model price of tradable and uh so basically you you you push on CPI forcing cost adjustment in sticky prices to re rebalance relative  
[00:35:27] prices. Okay. So, N tank now end tank uh end tomouth plus whatever. So, here if you look at these pictures there is no big difference. Okay. The  
[00:35:41] aggregate is almost the same. uh so uh I think this is always we look always thought about you know the financial authory versus a component of the economy that works in financial authority the antto-mouth and the other  
[00:35:55] guys so how do how can we think that this financial authory aspect is important well you need you need to sort of let the model generate possible um price adjustment and I would actually  
[00:36:09] um sort of play around with this new literature taking seriously that the elasticities are low in the short run and adjust in the long run. This is actually GBC disruption low elasticity now in the short run high elasticity in  
[00:36:24] in the long run. Lots of work by George Allesandria Julian Quadrini an old paper that we know very well is shameless elasticity 0.1 and he gets a lot of action to to to  
[00:36:38] one is really you know like basically I cannot live without gadget from China.  
[00:36:46] So to conclude must readad must think about project must read paper uh and I think there are two issues that could sort of make us think about monetary policy in the future. The first one is  
[00:36:58] to understand macronamics is shaped by the two-way feedbacks with theogenous dynamics in sectors and along the the the income distribution and along the the groups that may may not or may or may not participate in financial market.  
[00:37:13] This is an old question. I just push sector aspect in it because we have the relative price now coming in pretty pretty pretty pretty strong and more important what is the core inflation to target what is the proper stance right  
[00:37:26] in the previous paper there were layers of of of price rigidity so my simple natural rate does not work and actually with banko Canada we have now looking into the tariff again lots of work with  
[00:37:40] wholesale retail and production along the the the the the input output table to see how for example price adjustment may become more coordinated with large shock relative to price  
[00:37:55] adjustment less coordinated. So this I think is a beautiful new agenda ideally bringing forward more evidence and theory about the dynamic of trade fragmentation and I insist short and long run elasticities and I would add  
[00:38:08] one thing which I'm working now most of the time which is market structure those are markets populated by a few large uh firms trade trade disruption means oligopolistic change in the structure of  
[00:38:22] the market when you do bak kifari you don't have entry exit when you know all this kind of of of stuff that come from microsur to ignore this aspect and I think it's something we should think more about. Thank you very much.  
[00:38:36] >> Thank you very much Carlo. Great presentation and great discussion.  

## Q&A (00:38:42 – 00:47:56)

[00:38:42] So I still have one question but I want to first hear from the audience if there are any guido again. Let's see. I'm very curious >> monopolizing. No, but uh and I hope I'm  
[00:39:01] not stealing K's question. Okay. So, for the first time, let's uh I want to talk about the trans issue. So um when I think about um you know this  
[00:39:14] fragmentation no um I think about actually the there will be less international trade uh especially intermediate goods and therefore I want to be provocative that's I think is good  
[00:39:28] news at the end of the day for monetary policy let me explain why so I I I think the presenter mentioned some papers about determinance of trend inflation. I have a paper with Luca F  
[00:39:43] and J that shows that actually the China shock with China enters the WTO add a permanent uh deflationary pressures on in the US uh inflation  
[00:39:56] and um you know and monetary policy also was important of course inflation targeting but the other thing if you look at the cyclical instead uh impact is that Of  
[00:40:11] course uh then you have more of your marginal cost as a firm that coming from abroad and then immediately know there is some literature about you know the implication for the education Philips curve when you have um international  
[00:40:25] trade of course make inflation less controllable by monetary policy and simply because monetary policy affects inflation through the labor market and wages so and and domestic the price of domestic input right so if a lot of your  
[00:40:38] marginal cost is actually coming from you know India or China or whatever that means that actually monetary policy power to affect inflation is actually very very low and uh so actually I think  
[00:40:52] u I conclude that uh you know uh globalization is good news for monetary policy that actually China was u basically pushing out of business because they couldn't control inflation  
[00:41:04] any longer is fantastic for monetary post >> maybe back to Jenny and then we see  
[00:41:16] other questions. >> First of all, thank you John Carlo for the uh excellent discussion. Um I mean points taken uh it is true the only sector that we have  
[00:41:31] that is a sticky price sector is the non-tradables. Um CPI inflation does like targeting CPI inflation is suboptimal. Um but the we also have a tradable sector that is domestic and  
[00:41:44] which is not uh subject to sticky prices. So we would ideally just want to target the non-tradable sector. PPI goes a bit closer to that but it's still not ideal.  
[00:41:56] um the optimal allocations that we do so I didn't have time to present but we do uh look at uh what a social planner would do and there highlights a lot of the inefficiencies that J Carlo also mentioned in his discussion um that  
[00:42:10] there's too much reallocation from non-tradables to home tradeables there should be some in response to these import price shock there's actually too much because of stickiness in the non-tradable sector um there's a fall in real wages because of the adverse terms  
[00:42:23] of trade shock but also because of this um excess reallocation and this fall in labor demand. Um so some it's it's useful to do this optimal um efficient allocation exercise to highlight what are the the frictions in the model. It's  
[00:42:38] the sticky prices but it's also the intemporal adjustment of bonds. Um they yes so relative price shock I think that's really important to highlight that a lot of people think this is just a nominal a shock to the price level but it is a shock to relative prices which  
[00:42:53] has real effects on the domestic economy. So point taken to highlight that a bit more um on the elasticities uh yes um so we wanted to show that this effect is there without playing around  
[00:43:07] with the elasticity. So something neutral may be too neutral in your perspective. Um but it is plausible that for the front-loaded fragmentation scenario especially if we're talking about geopolitics you might want to target critical sectors in the other  
[00:43:22] economy. So yes you want to do something that has short you know target something that has short um in the short run has um low elasticities of substitution. So that's one motivation for that. Um and we have uh  
[00:43:36] different elasticities of substitution and extension. Um and we show how the effects in the gradual scenario could differ. Um but I think in the front-loaded case is the one that you would motivate most with different uh  
[00:43:50] smaller elasticity substitution in the long run. I think you know market forces would adjust. You would find other sources of supply such that it's a higher elasticities of substitution. Um  
[00:44:03] I have less to say on Guido's point but I point taken to to look take a look at that and yes I agree with you that it is um it has some benefits for the ability the transmission of monetary policy but I'll take a look at the the sources that  
[00:44:17] you mentioned seems helpful for us to reference as well. Thank you very much.  
[00:44:21] Um happy to take other questions bilaterally if there are no more here.  
[00:44:29] There's no more question from the audience then I take occasion to >> Yeah. So the the u I I thought that apart from the particular model that  
[00:44:41] sort of set results as a benchmark the questions there that they asked are very important particularly what kind of um indicators may be relevant what kind of uh in so in  
[00:44:55] in in the tariff debate there's been a big debate about whether is you want to abandon CPI targeting into more uh The other thing I want to say is that if you ask trade economist  
[00:45:08] they live with one problem which is the welfare relevance of this disruption is not that great.  
[00:45:16] So uh this is something uh important because it's sort of this rejoinder with the monetary policy and and this is opposite to what people think to what people feel what policy makers think to  
[00:45:29] what electional elections are won over. So there is something either wrong with the way we do it. Maybe we should have models that are more there is more action or more simply there is something  
[00:45:42] more to the trade disruption than simply the trade. I mean this is like a new regime with many many other things coming in and all this comes to your way. I mean I say that CPI impression target is not optimal in the model. You  
[00:45:57] tell people in the street they would scream at that right because the memory of the inflation in the past two three years we say you got to be kidding I want you to keep inflation down even if don't internalize the problem with activity so uh this is what I'm saying I  
[00:46:12] think we're walking into a new a new regime that uh will make uh our our certainties a little bit less solid than we had in the past. That's it.  
[00:46:28] Jenny or reaction then I can come back to my doubt I haven't yet understood um I don't know if I misunderstand the difference between the front-loaded and the progressive but if I understand in  
[00:46:42] both cases it's a permanent adjustment of tariffs but in one case it's frontloaded so it happens fast and in the other one it's long but in both cases the expected final increase is the same. So why  
[00:46:57] is the result then so different? Is it because you give space for this misallocations to to have a feedback effect? So thanks K. Um so to clarify in the gradual scenario that takes place  
[00:47:09] over a long period of time uh there's initial pullback in consumption because of the agents that do anticipate this and they see the whole path and this affects aggregate demands in the economy um which affects labor demands of  
[00:47:24] themselves but also of the other households. In the case where it happens immediately it doesn't matter if you can foresee it, you just adjust immediately.  
[00:47:33] Um and so this they they both result in a fall in domestic inflation. But also what matters for aggregate inflation is what happens to imported inflation. And in the case where it happens immediately, it's a larger spike. In the  
[00:47:46] case where it happens gradually, it's a smaller spike over time.  
[00:47:51] Thanks for clarifying. Thank you.