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.
=

# Internationalizing Like China

Authors:   
Discussant: None  
Video: https://www.youtube.com/watch?v=pNRINRlW4iw&t=0s


## Talk (00:00:00 – 00:20:59)

[00:00:00] internationalizing like china chris button is going to present you have 20 minutes great uh thank you so much and special thanks especially to uh for uh for coming to discuss this i'm really looking forward to hearing your  
[00:00:15] thoughts uh this is joint uh with amanda and jesse or both here hiding out in the back as well as with matteo uh so the starting point for this project is we've  
[00:00:27] seen a potentially a big shift in the global economy uh over the last couple decades which is the chinese bond market has grown to be the third largest domestic bond market in the world now  
[00:00:42] behind only the u.s and the consolidated euro area but until recently it was a market that was largely closed to foreign investment this has changed  
[00:00:54] substantially over the last 20 years uh and this potentially uh started as the kind of first stages of a fast internationalization process  
[00:01:07] uh with the potential goal of this of establishing the remember as a competitor or as possibly a displacer for the us dollar as an international currency  
[00:01:21] so what we want to do is we want to document a few facts about this international advertising process and then help to understand how this internationalization process has  
[00:01:36] so we'll start by presenting three simple stylized facts about the internationalization process uh the first is the rapid rise in foreign investment occurred initially among relatively  
[00:01:50] stable investors like central banks and then more recently has been driven by growth and private flows from kind of flightier investors like mutual funds  
[00:02:03] uh we will show uh some stylized evidence that this was the result of deliberate policy choices of the chinese government as well as that by the time the uh  
[00:02:16] the bond market had opened up to uh more fully to these flying investors the these remember bonds were actually looking more like uh kind of in  
[00:02:31] uh in terms of what they would be uh what how they were being traded like uh developed economy bonds rather than emerging market funds what we'll do then after just laying out  
[00:02:44] these stylized facts is develop a dynamic reputation model with imperfect financial intermediation in order to understand the behavior that the chinese government is pursuing in particular in  
[00:02:58] this model there's going to be a large open economy that is looking to raise external financing from international investors via bond issuance and we're going to  
[00:03:12] have investors who are worried that an opportunistic government might impose capital controls x post on them and make it hard for them to get their money out during a crisis what this is  
[00:03:26] going to lead to in our model is a buildup of reputation of the government as it maintains good behavior and doesn't impose external capital controls on outflows at low reputation levels the  
[00:03:40] government is only going to borrow from stable investors and build up its reputation as its reputation improves it's endogenously going to open up to the flight year investors uh international okay  
[00:03:55] so let me skip over lead review in the interest of time and i'm happy to come back to specific questions uh during the q a so the first fact i'm going to quickly document is the rise in foreign participation early  
[00:04:10] on this was largely concentrated among official investors particularly central banks and sovereign wealth funds so here what we do is we get the official  
[00:04:23] central bank data from the imf cofer and we look at chinese remember reserve holdings we can see especially as of 2016 this  
[00:04:34] was essentially the entire investment or the entire foreign investment in chinese bonds by 2020 there was a substantial growth in private  
[00:04:47] investment in the chinese bond market uh there's a few unknowns here i'll just give a quick uh kind of view of uh our best guess uh based on uh who's holding  
[00:05:01] uh the part that we aren't able to classify of kind of the composition-laden sample of the split between reserves and private investment this is happening for example because we have a lot of  
[00:05:16] foreign bond holdings uh in singapore but we don't see the currency composition of those plans so the second fact uh is that  
[00:05:28] that we want to kind of lay out here is that the participation of foreign investors was in part it to a deliberate policy choice of the chinese  
[00:05:40] so in particular what we show here is data on the first date that a foreign investor accesses the chinese bond market through one of a set of programs set up  
[00:05:55] by the chinese government for approval for access to the domestic law okay uh so early on in the sample these are kind of oh sorry uh so early on in the sample  
[00:06:07] these are fairly restrictive uh pro application processes with long lock in periods and quotas we see very little growth here and then over the early and mid 2010s especially with the  
[00:06:22] uh introduction of cp cibm direct where there was a a large opening of access of the bond market without quotas to long-term  
[00:06:36] investors so here this was defined as institutions like central banks and sovereign wealth funds we see a large increase in approve i'm sorry  
[00:06:49] we see a large increase in investors see as the first entry uh among official investors as their first entry through one of these programs what we see then is where the  
[00:07:04] private investment really starts to pick up is after the introduction of the bond connect program which allowed a much more broader and unrestricted access  
[00:07:16] to the chinese bond market uh through hot access through hong kong uh by the way sorry i forgot to mention this is a cdf so mechanically it has to sum to one by the end of the sample  
[00:07:31] uh the last fact i want to quickly establish is kind of a bit of quantity evidence that rmb bonds are actually look a  
[00:07:42] little bit more like dm bonds than uh like em bonds one thing i should quickly say is that uh which i should have mentioned earlier uh the vast majority of what  
[00:07:56] these investors are buying in the bond market are either the explicit government bonds or the policy bank bonds which are at least  
[00:08:10] seen as having implicit guarantees from the chinese government and so what we show here is if you are holding uh if this is among  
[00:08:22] mutual funds using morningstar data if you are holding bonds denominated in say u.s dollar currency at the left hand of the asset  
[00:08:37] at the left hand of the axis how much of your remaining portfolio tends to be in other dm bonds at this end high we go down we see  
[00:08:50] the japanese yen right next to china and then as we go down the spectrum we see a large set of em currencies so this is just a bit of stylized evidence that  
[00:09:04] china looks some like it's somewhere either at the bottom of the dms or kind of at the top of these a similar bit of evidence this is looking at if a mutual fund moves in to  
[00:09:18] chinese bonds what is it moving out of this is essentially a within fund portfolio shift uh what we see in 2019 it's essentially entirely coming out of u.s treasuries and other developed  
[00:09:32] economies uh in 2020 a lot of it is coming out of agencies and u.s treasuries so just a bit more of stylized evidence what are they moving  
[00:09:45] what are these uh mutual funds moving out of when they're moving into chinese bonds they're moving out of developed markets so now let me go into the model so the  
[00:09:57] way i'm going to set this up is go quickly through a dynamic financial intermediation sub game then that's going to be embedded into a dynamic reputation model that we're going to use to make sense of these stylized blocks  
[00:10:12] brief summary of what this model is going to look like so it's infinite horizon and discrete time each date is going to have a sub game where a large economy government is facing world investors  
[00:10:27] buying debt these governments can be committed or opportunistic the difference being committed governments would never dream of imposing a capital control whereas opportunistic governments will strategically choose  
[00:10:40] whether or not to impose capital the sub game is going to have kind of a familiar uh financial intermediation structure to it very much in the spirit of homestream tyrol type setups so they're going to  
[00:10:54] start by raising debt and undertaking long-term investment at the beginning in the middle a state of the economy is going to be realized there might be a binding  
[00:11:07] constraint and they might be forced to liquidate assets at this point opportunistic governments are going to be tempted to impose capital controls so what does the value of intermediation look like here so long-term  
[00:11:22] project that pays off at the end of the sub game liquidation discount in the middle of the sub game in the low state of the economy there is a collateral constraint where you have to roll over debt backed by collateral subject to a  
[00:11:37] haircut the temptation of the opportunistic government here is going to be that imposing a exposed capital control ends up relaxing the collateral constraint and allowing for more debt before  
[00:11:52] what this means is that the low state value is going to be a multiple of the liquidation value of the bank critically this uh net worth multiplier on the value of the  
[00:12:05] bank depends on the whether or not you impose the capital control there's also a high state which basically generates the same thing but the net worth multiplier is a fixed uh value that i  
[00:12:18] won't go too deep look into who are the investors we're raising money from so we're raising money money from stable and flighty investors big difference stable investors have a  
[00:12:31] smaller haircut but they're capacity constrained we set up a they have essentially a adjustment cost from investing into the country this is going to give rise to an upward sloping demand curve and the  
[00:12:46] interest rate uh the country is large so it takes this into account this interest rate schedule when it's choosing its issuance strategy if capital controls are imposed exposed  
[00:12:59] this lowers the exposed return so what do investors care about they care about this object m which is the probability that capital controls are imposed exposed we're going to call that the reputation  
[00:13:13] of the culture so just to give a brief look at kind of what the policy functions of the committed government look like as a function of its reputation so at low  
[00:13:25] reputation levels it's optimal to just borrow from stable investors below their capacity constraint debt financing is still still too expensive uh  
[00:13:39] in the interim you've gotten to a point where you would want in an unconstrained optimum to exceed their capacity but you're not able to uh at that point it's not worth opening up  
[00:13:52] to flighty investors yet because you have to bear the higher haircut eventually reputation would get good enough that you would want to actually open up to fly the investors and then you're back at an unconstrained  
[00:14:07] similarly the interest rate schedule is falling and then jumps up with the dash once so what is this all biased this all buys us basically that there's a flow and direct utility function in to  
[00:14:20] the committed type government and what is the uh opportunistic government if yet if they impose capital controls they get a proportional scaling of that flow indirect utility function this is now  
[00:14:34] what we're going to take feed into our dynamic reputation model so the key thing to note here is there's going to be type replenishment in the dynamic reputation model the government's type follows a markov  
[00:14:46] process uh where they sometimes die and get replaced by the opposite type uh this is very much to the kind of phalan amador failing tradition what it's going to end up yielding out is we're going to  
[00:15:01] get an equilibrium fixed during mixed strategies so what's our equilibrium going to look like we're going to have a t period cycle zero to capital t there's going to be an opening update where we let in the  
[00:15:15] flighty investors and there's going to be a graduation date where the opportunistic government plays mixed strategies and builds up reputation prior to the graduation date and then eventually the temptation to be  
[00:15:29] imposed capital controls becomes so large that the opportunistic government imposes them with probability one and resets to the beginning what's the basic trade-off of the  
[00:15:43] opportunistic government it's just if they don't impose capital controls they get the value of the committed government and they gain reputation if they impose capital controls they get a higher value that  
[00:15:56] period but they lose reputation uh so mixing requires the to be equal let me uh just say quickly about kind of how this ends up looking and what this structure  
[00:16:10] has managed to buy us is that basically when we go and put this all together what we get is essentially an indifference condition on the flow utility function we put that into a transition dynamic what do we get  
[00:16:25] out of it we get out an ar1 equation for the buildup of flow and direct utility or of reputation what's really interesting here is that you get a  
[00:16:37] point at the point of opening up where you get a jump in the transition dynamics of this equation what's that mean it basically means you get a particularly large buildup in indirect utility at the point the  
[00:16:52] economy opens up why is this it's basically because opening up is an expensive action for the opportunistic government because the opportunistic government loses essentially  
[00:17:05] uh as the haircut grows after you open up the porsche the net worth multiplier of the opportunistic government falls from imposing capital controls and so you get  
[00:17:19] less of a value from waiting until you open up to impose capital controls now let me spend the last two minutes uh sorry uh with only one investor type we can show there's a unique equilibrium of  
[00:17:32] this form uh i won't go into the second proposition uh we're still working on a little more on that one let me spend the last two minutes just uh talking about how we go we take this  
[00:17:46] uh model uh give a numerical example of it and see how it helps to explain the data so here we give a numerical example of a kind of simulation of this model what we see  
[00:18:00] happening here so here we have an opening update of two a graduation date of five what happens to a country that begins this process they start at a low reputation they slowly build it up up to  
[00:18:14] the graduation point how is this happening they start off by with a low probability of mimicking the probability of mimicking the committed type grows over time and then eventually that graduation  
[00:18:29] falls back to zero what happens to the policies on this equilibrium path so at low reputation it's not yet worth issuing to flight investors so we borrow  
[00:18:42] from the stable investors below their capacity as we build up we build up our debt issuance we hit the capacity of the stable investors and then we open up and we get a big jump in issuance  
[00:18:57] uh in debt issuance at the point of opening up because uh at this point we're moving from a constrained optimum to an unconstrained optimum what happens similarly to the interest rate schedule we get a big drop  
[00:19:11] in the interest rate schedule uh as we build up reputation it flatlines as we open up given we've got a huge jump in debt issuance decision and at that point actually it's interestingly at least in this case uh  
[00:19:25] by the time we've opened up we've gotten a lot of the uh total kind of drop in interest rate schedule and increase in equilibrium bonus once uh this is kind of you know the kind of  
[00:19:39] loose mapping to uh the idea that you're looking a little bit more like a dm currency so we've got a couple ongoing extensions on competition of two-way flows i don't have time for them so let  
[00:19:52] me just go ahead and work uh wrap up so what we've given here is just a simple theory of uh how uh how a country in particular in this case china is building up its reputation to  
[00:20:06] potentially become an international currency provider our theory is at least uh at a high level consistent with the internationalization strategy that china seems to be pursuing and it gives us a  
[00:20:20] lot of interesting directions uh we think to go in the future like thinking about competition uh between uh potential reserve currencies uh so thank you all very much i'm really looking  
[00:20:34] forward to hearing uh thoughts from uh olivia and from everyone else thanks chris the discussion is okay  

## Discussion (00:20:59 – 00:38:36)

[00:20:59] great so thanks a lot for inviting me to to discuss this so you know it's a great paper super creative and novel facts and models so you know it's really a new topic so i don't have much expertise on that so  
[00:21:13] so what i'm gonna do is mostly i'm going to talk about the kind of thought that this uh triggers you know what kind of stuff could you look at in the data and maybe you know add to the model so just to summarize you know the new facts or  
[00:21:26] this pattern that early in the 2000s you know you've had central banks serving wealth funds entering with restrictions so kind of excellent capital controls so these are the stable investors and after a while you know china opened up to the private sector to these flight investors  
[00:21:41] like mutual funds uh and after opening up your thanks to uh to those you got strong private inflows and there's this evidence that you know at the same time you saw these outflows from uh developed market bonds so  
[00:21:55] treasuries uh you know agencies and so that suggests that well it's kind of substitute they're not coming from very risky funds or risky corporate bonds or other emerging markets so i think the key question is why do we  
[00:22:09] see the sequential opening up the key question that the model is trying to answer and so the way they they answer that is that why it's really the optimal path for building reputation when you have these multiple investor classes so in the nurture the intuition is that  
[00:22:24] the flight investors they demand higher rates um so you don't want to let them in until you know you really have to because you've exhausted the stable investors capacity this d-bar  
[00:22:37] and this will and you will only do that also when your reputation is high enough because the rates that you have to pay are very high the reputation is going to lower the rate that you have to pay so as reputation goes up you know it becomes cheaper eventually it's worth it  
[00:22:52] uh and in in a way it's saying that you liberalize only when investors already treat your currency as relatively safe and so that comes out just to frame in the big context uh gita gopinat was saying this you know we need to go about  
[00:23:06] china if a country wants to become a global currency what do they have to do they have to liberalize uh have you know free uh free mobile capital uh anyway you know that's the conventional you that's the way to achieve that what the model is saying is  
[00:23:20] that well it's only worth it when you're already close to being viewed as a global currency so don't do it before in a way emerging market shouldn't do it because they have low reputation so it would be very costly uh there's a little bit of twist which is when you open up  
[00:23:34] you get the game so maybe there is something to the fact that opening up by itself has a benefit but you know it's only when your reputation is already high so it's not automatic it's not a causal effect in a way in general just  
[00:23:48] only when you know you should do it okay so it's a complicated model as you saw so i won't do the usual thing where i do a simple version of that because it's kind of impossible for the reputation thing you cannot do two three periods so i  
[00:24:01] just use a figure of their actual equation so the key constraint that you're trying to relax with the capital constraint is the first one where it's a binding criteria constraint in the bad state and so you want to relax does this allows you to roll over  
[00:24:15] more more debt you know limit the outflow and you see that on the numerator you have one minus h so that's kind of the idea that it's going to be more beneficial to impose a capital control when you have  
[00:24:28] a low age investor so the the stable sorry the the stable ones uh because suppose you know h is one so it's super flighty so in any case you're not going to worry about anything so imposing the capital control won't help you so that's  
[00:24:42] why you get eventually this reputation gain when you open up it's because the benefit from imposing the capital control is just lower when the haircut is really high so then you know if you do that you get  
[00:24:54] this jumping utility proportional this g that chris showed but the trade-off is that if you impose the capital control you also reveal that you're opportunistic you're not the committed type that will never do it and so that resets your reputation to a low level  
[00:25:09] uh so the benefit from reputation is that you can borrow a lot at the low rate okay so you know to to finish the model uh setup so in the one investor type case you have a very keen dynamics this ar1 process just looks like this so  
[00:25:23] reputation evolves so you know v is a value function but basically view this as reputation because it's increasing in reputation so on the left hand side you have the next pr reputation which has to be high for to prevent you  
[00:25:37] from defaulting now or making you indifferent and so that's the right hand side so that's how it increases over time until you hit the graduation date and the point of the graduation date is that reputation will be flat afterwards so you're paying very low interest rates  
[00:25:51] they think you know you're very good uh and maybe you know that's the case of many countries now and you know another felon they use this model to talk about why you know uh how does time since the last default explain the spread and that after a while if you haven't defaulted  
[00:26:05] for 30 years you're not gain gaining further lower interest rates so the idea is a bit like this here so the key point is your debt increases interest rates fall and so when you have those two investors  
[00:26:18] as i said you know you open up later because flight investors charge a higher rate so that's the cost of you know going for them first uh now the benefit of the capital control as i said is lower  
[00:26:32] once you have the flighty so that's where you get the boost and that partly you know justifies what people say you know you should open up but it's a very solid explanation it's not just like opening up by itself will make you a global currency uh you know if an  
[00:26:46] emerging market randomly opens up it's just going to be very bad there'll be stock with higher you know high interest rates and they maybe won't be able to impose capital controls or or there won't be a gain so they shouldn't do it in general  
[00:27:00] uh so you know let's start with the comments so i had just a few on there first maybe it would be nice to see a little bit more on this portfolio rebalancing this because the key point is that well you see a lot of outflows from the dm bonds  
[00:27:14] uh and so it's kind of a particular period so you to see whether we can extrapolate i'm sure you know we can but it would be nice to add maybe something about what's happening actually in these uh countries so maybe measure the qe exposure of these funds you know using  
[00:27:29] the soma holdings data that nina mentioned too uh and for the en part which is not important not that big here just because they're smaller i'd be nice to see whether actually you know because they the imf is saying oh no you shouldn't put prudential  
[00:27:44] excellent capital controls for emerging markets and so maybe you know now they have nowhere to go so they have to go to china which is opening up at the same time but i think the main thing is probably qe because the big you know as you see the big thing is in blue which  
[00:27:58] is you know coming from the us and europe uh the second thing is that the key inside one of the key insights in the reputation is that uh as we said as john also said you know you should never let a good crisis go to waste that's a key moment when you can  
[00:28:12] build up your reputation in good times nobody learns anything about your type so mapping calendar time to model time is a bit difficult you know these cycles were in terms of counting the number of crises but one interpretation of the  
[00:28:24] data would be okay we see the 2010s and the reputation in a way is growing really fast because you go very quickly from you know closed to open up to mutual funds uh to flight investors  
[00:28:37] uh now the question is well why did reputation increase so fast you know we didn't see that many crisis that let china you know show that it was a good uh currency and that imposed and in a way the big one that we saw you know  
[00:28:50] maybe it was 2015 where it's not clear they really did like that equals zero uh they kind of depreciate that many things happen so it's not clear that they're getting reputation but right after that they did open up so so it's a bit you know maybe it would be nice to discuss  
[00:29:04] the actual bad events first events and how to interpret china's reaction and how they led to an increasing reputation and kind of everybody's talking right now but maybe 2022 is the next one they depreciated a lot uh last week  
[00:29:17] uh the economy is weaker and so on so so let's see what happens then uh now going to the model part so thinking you know it's it's really nice uh you know to formalize this idea of sequential opening up  
[00:29:31] uh one thing you know in terms of clarity it's here you have both accented and exposed heterogeneity so in a way you know the flight investors are just really bad because they charge you a higher rate and they're gonna go away in the crisis  
[00:29:44] uh now you need that because you know the committed type doesn't care about uh whether they fly because the committed type will just never impose the capital control so they don't care about the haircuts we can't get it endogenously so i think it would be nice  
[00:29:58] to tweak a bit the model where maybe the only difference is flightiness and so if the committed types they also impose some capital control sometimes if the crisis is large enough uh then you know the re it's not as clear-cut you  
[00:30:12] know but you also get the splatter maybe that because of that uh you open up later to the flighty just because their flight is not because in addition you know they're they charge a higher excellent rate i think that would be useful too i'm going to come back to  
[00:30:26] this to think about maybe the trade-off is actually that they charge you a low rate and they're flighty so maybe in reality it's the opposite you know there are more risk they're less risk averse and they're more aggressive uh but they're flighty and so maybe what emerging markets other than china are  
[00:30:40] doing why aren't they doing the same thing it's because there's a trailer between the two excellent and exposed uh dimensions uh but i'll come back to that so so the other thing is you know here it's friend in terms of tau so i'm gonna do something really bad like a capital  
[00:30:55] control you won't be able to take out your money uh why it's helpful because it relaxes the criteria constraint as we saw and it's very clear you know you can say well they did a capital controller or not think something more intermediate  
[00:31:08] and very close to the model in fact is that you could just use conventional monetary policy to decrease the rate and this would have a very similar effect of you know increasing your pledgeable income so that's what foreign foreign  
[00:31:22] as the central banks put now it's not free you know it's not that deep you know decreasing your interest rate won't have reputational consequences why because that will imply you know from uip that your currency has to depreciate  
[00:31:37] to compensate with an expected appreciation the the new investors who come in during the crisis who you need to offset the outflows so in this context you know committed type is a bit less clear but you know maybe that's the main thing that they use you know monetary policy  
[00:31:52] exchange rate not hard capital controls which are very extreme and then if you use that it's like a default maybe it's a lot softer but then you know what's committed it's a bit less clear it's it's and it's interest rate peg or it's change rate back uh and so you know in  
[00:32:06] the data you see that uh actually in this period of stress like 2015 you had this big depreciation and then afterwards you so so it moves around a lot um so you know now it's not pegged anymore but  
[00:32:20] uh do you interpret this as tao you are going from zero to positive or yeah it's a bit on here um so i think it would be nice to expand the interpretation to you know exchange rate depreciation  
[00:32:33] um another comment related to that is that you know one very important thing for china is reserves uh and one question you know is do they complement or substitute reputation so one way you can view this figure which is showing you reserves over time  
[00:32:48] is that initially they accumulated a lot then they used up a lot during this 2015 episode uh so maybe they were trying actually not to depreciate too much uh so this was good and now it's kind of flat they're not calculating again  
[00:33:01] so yeah it's not clear how to interpret that is it that now you know we have achieved the global currency status or close enough so you know you don't need to mimic anymore because you're close to the graduation date so that's why i don't accumulate reserves so you know this  
[00:33:16] figure almost looks like the m the mimicking probability figure in the paper or is it that you don't need it now anymore because your global currency and so the model doesn't have those but it's a bit you know it's i think it's an important thing to talk about you know  
[00:33:29] uh if you're safe do you need reserves if people already think you're safe or is it you know is it that you you don't need to give you that safe anymore but one key yeah now people worry too because it didn't  
[00:33:41] recover to the 2014 levels before the previous kind of outflow and so if this happens again in 2022 uh we'll see so one thing what's specific about china you know i've talked about other years  
[00:33:55] so that's not really discussed in the paper but if you have twice three things like growth size strong government in a way every country should do the same thing you know it's not about china the fundamentals would just affect the details of you know uh how slow or where  
[00:34:10] when does the graduation date and the question is maybe show other countries that did this you know maybe it was too long ago but it would be nice to highlight this pattern for other countries than china in fact the typical pattern is kind of in the opposite direction you let the  
[00:34:25] flighty first and that's why i think the new fact on china is surprising because we're saying oh they opened up and you know but later to the friday and these are not coming from events so so it's kind of uh going  
[00:34:38] against the other emerging markets so how do we interpret that is it that other countries are making mistakes and maybe it's a private sector's mistakes so these are externalities because here we're taking in the model the country's perspective maybe you need strong  
[00:34:52] prudential excellent capital controls to implement this path or i think you know another possibility is this thing where accented and exposed heterogeneity go in opposite directions so you have more flight investors but they're  
[00:35:06] willing to charge you a lower rate exactly and so then you trade that off and maybe then other countries are doing the right thing given those different configurations uh okay so okay i'm almost done so  
[00:35:20] one thing you know going to the extreme when you think about reserve currencies uh it's here if you wanted the definition to be well you know usually we think if you issue more debt you have to pay higher rates and that's true for a given reputation  
[00:35:34] but what's happening here is that thanks to the endogenous increase in reputation on the path you see lower rates declining rates and higher debt and so maybe that you know reserve currency would be once you achieve that very low  
[00:35:48] rate in spite of a very high debt uh another important feature that may be hard to have here with just one country is capital inflows so in bad times you you appreciate so here the best thing you can do you know even the committed  
[00:36:02] type is that you're going to have a small outflow it's still going to be bad uh so maybe you need multiple countries to think about those and i know they're working on these extensions but so they mentioned this extension with quantum competition between different  
[00:36:16] issuers uh and so you compete to get the monopoly rents from reputation so i think the way maybe to formalize or what they have in mind is that the country i would pay a rate ri divided by this function m  
[00:36:31] which is itself like an aggregator or some kind of cs aggregator over my reputation and your reputation and then there's a issue of you know what kind of strategic complementarities we have between large countries between small and large or between small countries so a lot like  
[00:36:45] what we do in iowa with the oligopoly models uh but i think one thing that's interesting to think about too is that uh with multiple insurance you can think about this exposed competition so basically the flight investor they have  
[00:36:59] to go somewhere so they could retrench domestically so that's what you know alpena don't have a paper about retrenchment like that but they also fly to other safe currencies and the key questions here i think which you know the multiple country model  
[00:37:13] would address is so china is coming now but in a us-dominated world and so in general is it easier to build a reputation when you don't have other safe currencies or when you do so and it's not obvious because so suppose  
[00:37:27] there's no other reserve currency so we're all bad and i emerge and i'm the first so so the benefit is larger because the rents are higher okay from the substitution but exposed there's also more temptation to default impose the control because  
[00:37:41] where are they going to go you know and so it's unclear and in a way you know if we think of outside china what were the reserve currencies before you know so you had the uk and they mentioned this episode where in 1931 there was a big  
[00:37:55] depreciation of the pound and that was kind of when the uk lost its status so a big loss in reputation in the u.s in 71 you had a big depreciation too but you know maybe there was a smaller loss  
[00:38:10] and so why maybe while in in 1931 you also had the u.s emerging but in 71 who else are you gonna invest in uh and so that would make a big difference and so whether china you know the fact that they are emerging in the  
[00:38:23] dollar world uh makes it easier or harder it is i think it's a question um okay so so i'm not here and thanks a lot again for inviting me to discuss great thanks  

## Q&A (00:38:36 – 00:47:58)

[00:38:36] do you want to take some questions first i wanted to say thank you for the usual of doing a better job of  
[00:38:48] explaining the paper that i did um so i appreciate that you can respond to something like maybe the point about  
[00:39:02] i'll just ask the reputation um like you know is it really about reputation they really increased that much over the short period in the 2010s when there was a 2015 kind of blow up um  
[00:39:16] maybe it's not reputation maybe it's just like size or something and you know trying to grow a lot and the flighty guys if they're flighty they're not gonna come in until it's like deep enough liquid and up the market um so maybe it's something other  
[00:39:31] than reputation yeah so that's definitely something we'll need to think more about um you know part of that uh there are parts of that where we do think that um  
[00:39:45] there's kind of specific parts of china that are uh very much kind of uh kind of prerequisites to a country  
[00:39:59] building up to be using something like an international currency and from that perspective we think uh you know uh the fact that china's a large and grown and rapidly growing country with a  
[00:40:12] deep with an already deep bond market is essentially a prerequisite to it to becoming the international currency in the first place and so i  
[00:40:25] from that part of the perspective i think we would view them as complementary stories to each other i think where the reputation part comes in is really just the part of convincing the international investors that  
[00:40:40] we've got this deep bond market that if you trust us is kind of a deep pool for you to access and the reputation is just that  
[00:40:54] little added step of basically saying we have enough confidence in the chinese government uh that we will be able to actually use this as a treasury type substitute uh to to essentially be able to fully  
[00:41:11] capitalize on that part of the uh that part of the dynamics this is also part of where you know so for example one of the things very much brought off that i think is a very good criticism is we  
[00:41:27] have kind of the dual role of uh stable and flighty of stable being kind of the double lower haircut uh cheaper funding uh  
[00:41:41] this is a this is a another good i think uh critique of the way it's set up right now i and it's something that we need to um we're  
[00:41:54] currently thinking about relaxing one of the reasons we potentially think in that framework about a country like china wanting to access the stable guys first is because they're  
[00:42:09] they're able to build up a potentially large store of these guys conditional on the expectation that china may eventually become an international currency  
[00:42:22] in a way that say a lot of emerging markets would not necessarily be able to attract large source uh large stores from this  
[00:42:34] stable type investors so one one view of that would be basically the effective capacity constraint of the stable guys is much lower for  
[00:42:48] uh smaller countries than for a country like china and that would actually potentially in our model suggest that you know if you don't if you if you went  
[00:43:02] all the way to the limit where there was no stable capacity at all or just a tiny amount of it you would essentially open up almost immediately to the flighty investors just because  
[00:43:15] basically you need them to raise capital for your country sometimes i'm trying to wrestle with what are the fun what do you how do you think about the fundamental characteristics of stable versus flighty industries right what makes somebody flighty is it i think you modeled it as  
[00:43:30] just a preference difference but you know we all most investors like more money right so is it that they have less information is that they have different risk profile is it that they are more or less diversified and where i'm going with this is that i wonder whether things the  
[00:43:44] government is doing might change the flightiness aside from developing reputation but also act on some of those characteristics yeah so that's a great question like the one of the cheap ones that we think of  
[00:43:59] this is like the the difference between something like a mutual fund and a pension fund where you've got kind of very different liability structures of the two  
[00:44:12] institutions and so there we're thinking about it coming from essentially these are different institutions pursuing different business models with different investment horizons  
[00:44:24] and in that sense it's something like the mutual fund is doing something like a more active monitoring of its investment than the  
[00:44:37] pension fund is and so it's going to be respond more aggressively uh in its response uh and so the way we founded this here was just they've got different haircuts  
[00:44:51] basically and so you need to post more collateral to convince the kind of unstable funding mutual fund to stay in the to stay invested in the country than  
[00:45:06] the kind of more stable funding pension fund or insurance company but it's very it's very much uh you know there's a haircut they demand uh and we  
[00:45:20] don't right now do very much to micro found here is where this haircut comes from yeah i have a question i wondering why this applies specifically to china in the sense that you know  
[00:45:35] basically any i mean you just do talk about like any emerging market countries could apply this and so are you saying that basically china just did the timing of this opening of capital markets just the right way so in a sense it's like a  
[00:45:49] normative paper and you're saying china did it well congrats or is it more like descriptive of why did china did it this way so i would use as more descriptive of china um part of that  
[00:46:04] reason i would not um i would not view this as a critique of the strategy of emerging markets um part of the reason for that is just  
[00:46:18] uh you know our our leading at least or my leading i won't speak for my co-authors um my leading kind of uh explanation is that uh  
[00:46:32] the longer term investors are sort of willing to put more money into the country to begin with when they think it has a chance of becoming a kind of  
[00:46:45] long-term term safe store of value with kind of a deep pool and that type of characteristic may come easier to a country with a deep  
[00:46:58] capital or a deep existing bond market than to a smaller emerging market and so it's to say you know if you took that literally and said the kind of stable capacity  
[00:47:12] available for emerging markets is much smaller than for a country like china it would our model would suggest an emerging market should potentially even start up just  
[00:47:26] open directly to fly the investors or at the least should open up much more quickly so i i would not uh i would not uh and my co-authors are welcome to  
[00:47:39] disagree i would not view this as a critique of the strategy of other countries okay let's just take our coffee break and we'll be  
[00:47:51] back at 3 10. thank you