 Hello everyone and welcome to the eighth meeting of our virtual seminar series on central banking and digital currencies. Today's meeting is hosted by the Federal Reserve Bank of Philadelphia and we're very pleased to have Daniel Sanchez as our moderator Daniel over to you. Good morning everybody. Welcome to the seminar. So our speaker today is Marcus Grunemeyer, the Edwards S. Sanford Professor of Economics at Princeton University and Director of the Ben Hunt Center for Finance at Princeton University. Marcus Research publications cover the fields of financial economics, macroeconomics, banking theory and asset pricing. His vast publication record includes seminal contributions in all these areas. Perhaps an easy way to illustrate how influential his publications have been in the profession is simply to mention that as of yesterday, Marcus had over 35,000 Google scholar citations recently has published a book entitled the Resilient Society, which feature among the best, the best books of 2021 according to the financial times. Marcus has also been engaged in the study of digital currencies and other digital assets today's paper is platforms tokens and interoperability joint with joint work with Jonathan Payne. Our discussion today is Charles Conn, Professor Emeritus at the University of Illinois to vanish in pain and a research fellow at the Federal Reserve Bank of St. Louis. Charles has a vast publication record including works in financial economics monetary theory, banking theory and economics of payments. All of his papers are considered seminal contributions to these fields. And let us not forget that Charles has also followed a well known paper on a solution method to linear difference models on the rational expectations. And for all these contributions we are in his debt. Without further ado, I invite Marcus to start his presentation. Thanks, Daniel and thanks to everybody and that's as Daniel mentioned that's trying to work with Jonathan Payne who is also here and you can ask all the cut off questions to Jonathan. And so we will try to do to understand you know platforms and platforms which issue tokens and what role interoperability plays and also in connection to the theme of the seminar series, the role of CBDC. So essentially some tech trends essentially we have all these digital platforms and the digital platforms they're all about matching technology so if you're part of a platform you matched more easily. And at the same time these platforms also issues on digital tokens that's about the payment technology. And there's an increasing interaction we would like to understand the interaction between the digital platform to tokens and what role then if the center bank is providing some digital currency, like digital euro or digital in the other currency. What role does it play and how does it affect the interaction across platforms and also with the public market space. So the policy question is use how to regulate the competition between the public market and you know with the notice, you know the currency is the dollar or it might be CBDC or digital dollar and interaction with the platform there's an incumbent platform, and then there's also an additional potential with our internet platform and the platform is issuing a stable coin in form of a token so I use this as my token sign. And there's a competing potentially entered platform, which offers a different private token, which I call with this token sign or whatever the sign is with a prime. Okay, so that's the notation. I'm using for this talk. So the question we would like to address is how to regulate the competition between the platforms and also with the public market, and we should see the CBDC be a legal tender. So what's what's the role of CBDC as legal tender or not and what are the trade offs if you make CBDC a legal tender. And the key concept here will be interoperability so we try to understand interoperability and how interoperability is affecting this competition. So interoperability in one contribution of this paper is to unpack essentially the concept of interoperability, and we actually highlight three different ways to think about interoperability or three different forms of interoperability. The first one is exchangeability of tokens. Can you swap one token for another token or for the CBDC or without any fee. So that's what we call a token exchangeability. The second concept of interoperability is acceptability tokens accepted on all the platforms so exchangeability could still be that you have to use on a platform this particular token of this platform but you can easily swap it. If you have acceptability, that means you can actually even use some other token or the CBDC on any platform. And that's when CBDC is a legal tender. So the platforms are required to accept or members who are active on this platform are required to accept CBDC as a legal tender. And the third concept of interoperability is ledger portability. So if a new platform comes in an interim platform can you take over the incumbent ledger. Are these ledgers sufficiently interoperable that the new incumbent platform can take on the old platform or you know that can communicate accordingly with that. So these are the three concepts of interoperability I would like to discuss in a formal framework. And then we show essentially the effects of the exchangeability, they avoid this lock-in effect. So there will be a lock-in effect I will describe. And whenever there's a lock-in effect, the incumbent platform can charge a higher markup if you want to trade on this platform. So charge some fee for making some retail purchase and sales. And if you have exchangeability, but then you can't charge this markup the same degrees because there's no lock-in effect. This token acceptability destroys essentially the commitment via smart contracts. So we cover here the smart contract space where if you use the tokens and if you force to use this token, you can actually platform can offer some credit because it has additional commitment power towards the people who are on this platform. And if you allow CBDC as legal tender, if you allow acceptability, you destroy some of this commitment power, hence your fewer smart contracts, your less credit provision in this thing. The third result essentially is that when you have ledger portability, the new platform can't credibly inflict defaults on the old platform. I will be more specific on that, but there will be the incumbent platform is granting credit. And if a new platform comes in and everybody can switch over to the new platform, you can just default on the old platform. And that's actually a threat, a new platform, a new entrant has towards an incumbent platform if there's no portability. If there's portability, then it is the case that a new platform cannot really commit nor to take the credit commitments on to the new platform, to the entrant platform as well, so that the threat is less powerful and less credible. So this I would like to outline in a formal model. And I would like to enlarge essentially usually we use some matching models with a certain matching intensity. And there are no tokens attached to it. And the agents, because there's only a single platform, there is no platform choice for the agent. So we have to generalize the model framework quite a bit. So what we here do is we make a strategic, the platform is strategic, competes with the public market and with an entrant platform. And the platform is also issuing tokens. And the choices essentially platform makes on the retail space it changes the markup if you want to trade on this platform, you can charge a markup so the price will be plus this percentage markup psi. And then on the payment space, you can determine the interoperability of the tokens that can be an exit fee. So if you want to swap the tokens for dollars, you can charge something if you want to swap the tokens of your incoming platform for the entrant platform, there might be an exit fee and so forth. And then of course, when the platform grants credit, there's also an interest rate or some conditions on what conditions you grant is credit. The agents are buyers and sellers. So this neutral very simple data time reference rate of raw, and they have a discrete choice so we use some discrete choice modeling tools from international trade and IO, where we essentially said they have to make a choice whether which platform they go to the private platform or the public marketplace. So that's what what they're deciding essentially on. Let me be a little bit more specific about the model so here's the platforms platforms you have a public platform, we have a private platform then come in private which issues a token the public uses a dollar. And then there's a potential entrant, which has its own token, and the potential entrance come with a certain rivalry at lambda e. Okay, so they're not always the occasion there's a draw and then an entrant could come in. Then there's certain matching technology. So there's some, you come become a seller let's say with a certain arrival rate. So that's your hybrid lambda for seller. If you trade on the dollar platform or the public platform, and it has typically a common component of platform specific arrival rate. And then it has an interesting graphic for individual eye. Okay, so that's the SC so the product of the common component plus the individual component that's essentially for this individual eye that's the arrival rate with which he can sell something and find somebody to create with. What's important is that so that's where this international trade. So this see is a fresh distribution it's drawn from a fresh distribution and signals how much this person likes this particular public platform, or if you know he likes this particular private platform, which is this, you know, with a token sign here. So everybody gets some draws, and then makes a decision how much do I like the public platform or the private platform. And that's, you know, there has a common component to it and has an idiosyncratic component to it. And that's drawn and that's individually specific with ago some people really like to be on the public platform others really like to be on the private platform, and there's an elasticity of this coming out of this and the fresh distribution makes this discrete choice almost like a CS similarity is a very strong analogy between this discrete choice in the elasticity is to a CS what we typically have with a continuum or choice of a continuum but here we have a discrete choice only. And then you have similar for the bias with sort of elaborate of the bias, and so forth. So on top of it, occasionally, there might be new entrant platform as well. The new entrant platform could have a higher matching rate so it comes in with the matching edge, or has some bigger advantage compared to the other so when an entrant enters the public market with matching rate. So typically we assume when a new entrant would enter the public market place will also get this extra edge that the ratio stays the same says they saw very stationary. The timing of the agents, it's a continuous time setting. The agents observes this trading rate on the platform so this is lambdas that the agent decides which platform to search on, then the trading opportunities arise. And then all agents trade competitive at a platform specific price for the seller it's it's that price and for the buyer, it is a lower price and higher price because there's this markup on top of it. So that's roughly that's a setup to get the idea how the thing it's a search model it's with two platforms potentially an entrant platform there's competition there as well it's very much model like a food and bacteria taking this platform model and put this all together in this framework. So what we want to do. Now, use this model and use first a model without credit, and then a model with credit. So the first model, it will be the seller comes in, he sells something. And then he gets some tokens, and until he has to wait until randomly, he becomes a buyer and wants to buy something. So that's a model without credit so that's, we essentially everybody comes in as a seller first, and then get some token or some dollars and he prefers which platform he prefers. And then he can switch if he wants to add an exit fee, and then it becomes a buyer. Later, then I will do a model with credit and smart contracts were, you know, first you come in as buyer so you need to borrow something from a platform, and the platform has gives you the platform has more commitment power. Until you borrow until you become a seller because with random random arrival rate, you become a seller later on. So let's first go to this model one without the credit. Okay, so the platform is setting the markups for the good prices so this psi, and is setting here essentially the exit fee. If you want to move from the tokens into the dollar, or from the token to the new entrant token. So these are the fees the platform is setting and that's where the interoperability, the first interoperability when you set this equal to zero, then there's exchange interoperability. Okay, there are two decisions. There's a decision for each agent to decide should I use tokens or should I use dollars and occasionally when a new entrant comes in after the site should I switch from the old tokens to the new tokens to the new entrant platform. So what happens, you know, you are a seller. And, and then at the random point of time so that's what this basura I will get indicates become a buyer. Okay, so that's, and you have to decide then you know when you're seller, which platform to go to, and then which platform to do to search for and what whether all tokens or dollars. And then of course there will be other some other guys who were sellers beforehand, and then there will be buyers when you are seller you the others will be buyer so this is for the others now. And then, you know, there might be more buyers later on coming on as well. So the first thing I would like to illustrate is that if there is an exit fee you might say there might be a lock in effect already between the token and the dollar but it's actually not so obvious and it's there's not necessarily a lock in effect. Let me illustrate. So that's supposed to as an exit fee. But then actually, if you, if you are seller, you know, then it's good to have some tokens. So if the others. So this is the others, the potential buyers if they have a lot of tokens out there. So that's a fraction of wealth, which is in token this eater and dollar is the fraction of wealth which is dealt by the others in dollars. So if if you are seller, it's actually good to have some that you have some token. And when you go on but then later on it's actually the other way around because it's good to have dollars because you're competing with the other buyers. Okay, so first you you're on the other side of the market. Then you want to, you know, take some token but then later if the others hold it so if the others hold a lot of tokens it's actually also good to have some lot of tokens because there are a lot of guys out there, which you can trade with. But later when you're a buyer, you probably you're also competing with other buyers. So you want to have the opposite. So and everybody anticipates that that you want to have first it's good to be in the token platform and then switch to the dollar platform. But everybody anticipates that and essentially the exchange in order to lure you in onto the platform to lure you in. And it has to compensate the guys to come in in in on its platform for the future exchange fee you have to pay so anticipate this lock in effect. And hence the lock in effect is not so strong because you have to exchange or the platform has to compensate if to lure them in. Okay, so that's that makes that we can see the anticipation we can say lock in effect it might not even be there depends on the elasticities on this for shady distribution things. Well, there's still a lock in effect or even an anti lock in effect. Now, the situation is very different between the competition between the token there in incumbent platform and a potential entrant platform, which come in occasionally with a rivalry at lambda E. And it's very different. And why is it very different, because it's essentially the past sellers, they had no chance to hold this new in entrant token they have only the incumbent token. So they don't have this new entrant token because new entrant is only entering at this point, potentially when you make the decision. So there will be no past sellers to hold the new incumbent tokens, and hence that's very different. So now it's the opposite you have this lock in effect, the past sellers had actually no opportunity to hold this new entrant token. So there are no people out there have this they could actually pee with that and be on this new entrant platform. And that's why there's a strong lock in effect when you have a competition between the entrant platform and the incumbent platform. And the big difference is that with the public marketplace, there will be some other sellers out there and buyers who have the dollar already, and there's competition with that. You don't have this in this case, where, you know, the entire platform has no base yet so essentially you're born into there's a big effect of being on this incumbent token so that the exit fee is actually weakening then the possibility for an entrant to come in for an entrant platform to come in. Now let me go look at this case so that's a more interesting case for in terms of competition across platforms. There will be this lock in effect because as I mentioned before. So that limits competition across retail platforms. So what happens essentially. And let's suppose there's no exit fee from going from tokens to dollars that's not the focus here. And let's suppose the new platform is better is is a better has a better matching technology than the old platform. So the platform is setting so the token to token exit fee will be set at a maximum they charge huge 100% exit fees or in this stylized equilibrium. And they will also charge because it's very because of this lock in effect, they can also charge a high markup on the retail space. What they do on their token space they will charge high exit fees and on the retail space, but they actually charge a high markup that's given by the solution here so that's nicely solved and is a closed form for this particular case, where we can solve that and you see, that's the competitive edge of the new platform if the new platform is more competitive and has a better matching technology, then they cannot charge such a higher. So if this capital lambda goes up, this psi is not so high. Similarly, it was the share of the wealth which is in the token, if this is higher than the size actually going up as well. So that's, this is the competitive study you get out of this analysis. So that's the first thing is you have this token lock in effect that limits really competition across the incoming platform and potential entrant platforms. And that's why the incoming platform wants to charge a high maximum and low markup. Now, another special case to analyze, let's suppose the incoming platform, the entrant platform has no competitive edge. It's as good as the other one. And so you can impose interoperability, then you just impose, let's say no exit fees through CVDC as legal tender, for example, and then you actually will know markups that would be very competitive. So exchange interoperability, if you post exchange interoperability, perhaps via CVDC as legal tender, then you get very strong competitive forces. Now, let me now move on to the second setting where we now have credit, not only, you know, holding some tokens as an asset, but as credit. So here I focus on the case where there's matching platforms as before, let's suppose there are no markups, but the platforms can lend via smart contacts. So they are also setting the lending conditions. And that's, you know, through a matching we think like eBay. So it's not Amazon where you have to sell everything to Amazon, but it's more than eBay structure. And we have a production here as well. So you first a buyer now first, in the previous model was first a seller and then a buyer now you first a buyer you want to borrow something and then it becomes a seller, and you're producing something and then you buy one input unit and you can set knowledge than one output unit. So that's why you want to borrow in order to do something. And you still have this competition with a token and the dollar platform, and you can enforce essentially the credit differently. So that's enforceability. So if you're in the output, if you sell it on the token platform, you meet somebody who can buy your stuff at this rate. You know, this is freshie things. And the agents pay off is this and the platform affection of your pay off what you get you have to give to the platform. So that's the credit contract. And if you sell it the dollar platform, you know, you get the agents pay off is only gamma of this, and the platform gets zero. And what you can do is you can of course the default so you borrow from the official platform, and then promise, you know, when you sell you will sell it for the tokens and the platform can immediately take it away from you or you can just switch to the marketplace to the official public marketplace and collect the dollars and run away with it. And that's, you can see that this is the interoperability here. The acceptability makes already a difference where it makes it harder for the platform to enforce the debt contract. So without acceptability, you only default and sell at the dollar marketplace. So your personal, you know, this difference is small. Okay, so for example, the advantage of being of this more high matching rate of this token platform for you is is compared to the public platform is small with acceptability. For example, if CBDC is legal tender, but then it actually kills the commitment of the smart contracts. So it's actually impossible for the platform to enforce it. So there's essentially the credit the intermediary their platform can give and what happens then you get a whole different structure in the retail market space. So any matching platform cannot really implement this lending, because the CBDC is legal tender is killing that. So you will have something like Amazon's where all the trades have to go through Amazon. And it's not only a matching platform where the buyer seller meet and trade, which has to be always bought and sold by Amazon, some intermediary. So whatever you do on the CBDC side, on the acceptability side has huge implications on the retail market. Four minutes Marcus, please. Okay. So you also have of course, competition, coming for the potential entrant. So that's, again, the agents also choose between the token and the potential entrant token as before. And here's something interesting happens. There's a so called, we call this an anti lock in effect. So it actually hurts potentially the incumbent platform. So there is compared to the entrant. And what can happen is the entrant platform goes to the agents who have tokens or have credit with the incumbent platform say, if you come over to you, you can default on your incumbent platform. Okay, so why don't you come over. And that actually makes it very tricky. Let's suppose there's a new, a new entry coming as no competitive edge at all. It just comes to this agents and says, Okay, if you come over you can default on the old platform and just you have the same matching with us, why don't you come over and just default on your loans. And the incumbent platform essentially wants to fend this off and wants to pay off the entrant through potentially killer acquisition. So you're any whenever with lambda e arrival and new entrant might come in the old platform says, you know, I pay you off I acquire and make a killer acquisition, and then to buy off the threat and of course this creates some additional costs for the incoming platform. So it's the incoming platforms constantly when new entrants come in potentially come in, they have to acquire them and pay off this potential entrant, and that increases the cost of providing loans because you know, you have to pay off all this killer, I have to do all this killer acquisitions with there's no value created to that. The only thing it does, you just make sure there's nobody else coming in. And that means that you give a lower the loan book will be smaller, because the higher the loan book is the more you have to pay off the friend of this killer to make a killer acquisition the friend of potential entrant. And there's an interesting element to it. If your ledger is very portable to the new entrant, but then you can actually the threat of the entrant is actually much lower so if, because then actually the incumbent platform will say, you know, you will not really get this customer who owe me some money because when you take over my ledger, you will still go to the customers that please pay back your loan. So if there's a huge portability, if the portability works well between incumbent ledgers and the entrant ledgers. The entrant cannot really threaten anymore to say, okay, when I come in, I will, you know, take all your, all the guys who owe you something away, and then default on your, I will can say as an incumbent, no, no, you will not do this, actually, because you will just take on the ledger and will also still ask these agents to pay back their loans. So if there's portability interoperability, this threat is much more minimized. I mean, it's a very different form of interoperability, because it's about the ledger how technologically you can integrate the ledgers between the two. So that's, let me conclude with that. So what we try to do here is we would like to build a search type model where there's two platforms or potentially three platforms competing and try to understand how it interacts with the token issues. There should be, you know, retail platforms and payments, should they be allowed to merge that's an extra question I didn't address yet but the main focus here was how to regulate the competition across platforms, and what role does CVDC play. And there are different interoperability concepts, and some of them can be implemented via CVDC as legal tender, but there's this exchange interoperability this acceptability interoperability and that's at the end this portability this ledger interoperability. CVDC as legal tender, and this thing essentially restores competition between the private platforms. So it leads to lower markups on the retail platform, but it hurts credit provisioning via smart contracts because it kills this commitment problem. And this portability of the ledger reduces the necessity to have killer acquisitions all the time, and hence reduces the incumbent platforms to run this platform efficiently. But otherwise there will be excessive entry of platforms for excessive killer acquisitions in a sense. So that's in a nutshell, what the paper does is all about. Thanks again for giving us this opportunity. Thank you very much for what you said. Thank you so much. Now, Charles, you want to start your presentation. And you see the screen. And yes. Okay, five I think at five I will try to get if it's viewable at all we'll just go with it. I can see well. Okay. Thanks for inviting me to discuss Marcus and Jonathan's paper. Interoperability is the theme of the paper. And I want to start by talking about the importance of the issue. And then I'm going to spend some time contrasting the typology of interoperability that they use with in their analysis with the alternative typology that's currently in use, which I think is going to shed some light on the limitations and strengths of their model. And finally I'm going to give my interpretation of their model and of its conclusions. For quite a while, I've argued that the main issue for any potential new payments arrangement that hopes to get mainstream is how it will integrate with existing payments arrangements. How do you get funds into and out of it? How do you link up to someone who's not already part of that system? I think the belief in the centrality of interoperability is now conventional wisdom. And meanwhile, the crucial regulatory questions are what standards to set regarding that interoperability. Tech types fight lots of battles over the definition of interoperability, but we can go with the CPMI definition. It doesn't matter what's under the hood, so long as the effect is that A can just pay being, even if A and B don't get their services from the same payments providers. And that sounds like a good thing. And why doesn't it happen smoothly? Well, first of all, there are legitimate fears that interoperability causes operational risk. But then in addition, the platform literature notes that linking with rival platforms may reduce my market power. And so flattening out the demand curves for my services even while they're increasing the value of those services to my customers. And this fear is going to be reinforced if network externalities are important part of the story. The regulatory solutions to this problem, or the standard ones, requires some minimal degree of interoperability in return for meeting the safety standards. Now, Francisco and I are working on a paper right now in which interoperability is further encouraged by provision of public services, which while more basic than the private service, still reduce the importance of the network externality. So these kinds of arguments are applicable to all platforms. And in the case of payment services, Marcus and Jonathan want to unpack the whole story a little bit. And so they distinguish these three aspects of interoperability. They describe exchangeability, that is, no charges are imposed on redeeming tokens. Acceptability, that means the ability to use the tokens on other platforms. And finally, portability, that means the ability to transfer your account to another platform. In contrast, this policy work on payments concentrates on three modes of achieving operability. And let me go through them in detail. I apologize for the terminology is extremely unimaginative, but I didn't make it up honest. First of all, there's scheme interoperability. Now, scheme interoperability means that the different payments providers work within a single overarching open loop system. Think about check clearing within the Fed system. In the context of the current paper, scheme interoperability arises when all of the systems that are in the story accept CBDC, and all of the system tokens are convertible into CBDC. The second mode is called network interoperability. Two payments arrange schemes directly negotiate a mutual recognition arrangement. In international context, something like this happens when you're talking about cross-border arrangements for allowing customers to make credit card purchases. It's rarer in domestic markets and for the reasons we just talked about. And it's usually very complicated to arrange. So in the paper at hand, a simple form of network interoperability shows up in the final section of paper. So one mode is for service providers all to make an agreement among themselves. Another mode is to have an overarching kind of system that they all join. The third mode is called interoperability of systems, and the idea is this. Third parties appear and they act as bridges between the systems, developing the connections on their own by operating on both systems and thereby linking the agents who only operate on one of them. Now the classic example of this sort of thing is a traditional financial arrangement known as Huala, which transfers money by making credit and debit in separate unconnected locales. And possibilities like this need to be considered more in the framework that we're talking about here. So let me give my interpretation of the framework. We have competition between an existing public and private payment service with some imperfect substitution. There's a potential competition for the private service from a second private service, which is more or less a perfect substitute. Customers choose their platforms and then they go on to the platform to search for matching trade. And that arrangement rules out the possibility of multi-homing at the instant of trade. There's enough randomness built in that individuals could prefer to switch platforms when their status changes. The private platform is able to assess a markup on each payment made on the platform. It's also able to make a distinct charge for withdrawal of funds, and in principle it's able to do a charge or credit for joining the system in the first place. Customers value the ability to withdraw funds from the system, so they would be willing to pay up front for the ability to withdraw without charge. Now, if we consider competition between two such private payment systems, otherwise perfect substitutes, and if all the firms made a joint decision as which one to patronize, markups would be driven to a competitive level. However, an incumbent can put in place a charge to leave the system. For example, the charge could be in return for an upfront subsidy to join the system in the first place. This charge locks in the users, and by locking in the users like that, you increase the differential that an agent would have to pay to lure them away. If you want to think about it, you can think about this as exactly what a mobile phone plan does, give away the phones upfront and lock customers into contracts. Same story. Now, if we consider, notice though, that there are different charges for using the token for payment versus redeeming the tokens. This assumes though that the operator can distinguish a sale of goods from a trade of tokens. So a surreptitious Huala, and after all, Hualas are often sidelines to other businesses, could disrupt the arrangement, making a disguised arrangement that looks like it's going to be a sale, but it's in fact an exchange of tokens with somebody else, and therefore limit the cost of the exchange of a sales markup. Now, the basic model assumes that the agents get tokens by receiving them from others and in return for sales of goods. The extension model wasn't complete in the version I received, but the outline was clear. It assumes that the initial cash of tokens can be acquired by a loan from the payments operator. To ensure the repayment, you're going to need some collateral, and in this model the inventory of that is the collateral itself. As soon as the inventory has been sold, the assumption is that the payment arrives through the payment system and then a smart contract automatically diverts the appropriate amount to the payment system operator to repay his loan. That's an interesting twist on the traditional account of why banks integrate payment services and lenders. The standard story is that by observing the customer's payments and receipts, a bank can keep an eye on the customer's credit worthiness. Here, the argument is flipped. By watching when the money comes in, the payment operator can grab what he's owed before the borrower can withdraw it. This is reminiscent of processes that work in papers on clearing and settlement and wholesale tiering, and it may be more appropriate in that context than in the context they're looking at here, because for the story to work here, the payment system must be able to monitor all of the transactions of the borrower. If a sale can be made on another channel without the payment system observing it, then the power to collect evaporates. Doing it more generally, this is really an example of a moral hazard problem with side contracting or with side trades, and that's another old literature that probably ought to be referred to. The payment system's ability to collect on the debts is going to be limited to an amount equal to the convenience advantage that the borrower gets from using the monitored system, rather than going off and trying to use some other unmonitored system. And except in extreme cases, and by extreme case, I mean something like the story of a company store in a remote mining community, which can give credit in the assurance that the individual can never trade anywhere else. Except for extreme cases like that, I think this is a possibility unlikely to be significant in the real world. The third effect that the paper considers is non-portability. The sellers value the payment services, therefore they're willing to maintain the relationship by continuing to pay the initial loan. Sellers value the relationship because there's no alternative available. The lender will pay to maintain that power over payments, since without it his loans are worthless. Therefore, it's going to waste money keeping out entrance. However, suppose it's the case that suddenly a new powerful payments provider arrives. The former payment system would be able to negotiate with the replacement to obtain value from the transferred portfolio. And this means the incumbent will waste fewer resources trying to keep entrance out. An analogy is probably going to make that clearer. Customers of a mobile phone company have borrowed to buy lock phones from the company. A new mobile company enters and customers are going to switch into fault. Determined is the cost of doing business and reduces the incumbent's willingness to provide the services. So let the incumbent offer to sell its portfolio, loans and lock codes on the phones to the entrant. Certainly eliminates the wastage of the phones and it may maintain the enforceability of the contracts. But it requires a lot of steps. First, the new entrant has to be a monopolist in turn. If power is shared, the threat is weakened. And secondly, the new entrant must be able to identify the existing customers to hold them to the old debts and refuse them new service if the debts aren't honored. If these guys are able to sneak in as new customers, then the arrangement is going to fail unless the value of the phone to the customer exceeds the loan balance but they weren't a debtor in the first place. Alright, let me summarize. This is a brilliant framework for dealing with payments issues of fundamental importance. The basic structure used in the first part of the paper is surprisingly tractable and it can be used to get to real insights about the interoperability in payments. The second part of the paper, linking credit payments and smart contracts, is insightful and the consequences of it are worth exploring. My main reservation about the second part of the paper is that because results are so tightly tied to the model specifics they can't be used to address some of the key costs and benefits of interoperability. Thanks. Thank you. Thank you, Charlotte. Thank you so much. Marcus, do you have a few remarks you want to say? Yes. Thanks a lot, Charlie. This is fantastic. I think it really shows your expertise in this area. We're very grateful for your reaction and comments. So let me just make a few points on that. Jonathan can chime in as well. We haven't really looked at this Havala structure. That's an interesting thing we really have to think about. The default, let me say, you can actually default in our setting and move to the other platform, but you want to stick to the old platform if you have a higher matching rate in the old platform. So there are some guys who get idiosyncratic fragility shocks and they think they will find much more quickly a counterparty to trade with if they stay with that. And that's why they're sticking with the tokens. And others who have a lower matching rate, they will actually default. So there is the possibility to go to the other platform, and there will be a fraction of the guys who always move over there too. But all the other things are fantastic to put in perspective, and you have a much broader perspective here on the interoperability aspects. So we very much appreciate it. We'll probably follow up on you to get all the related literature and this framework as well. Jonathan, do you want to chime in as well? Yeah, sure. You're right about the default ability. So to default, you have to go away from the platform and trade. So that's the cost, as you say. So yes, if you're the only store in town, then you can enforce the contract very well because there's really nowhere else to go and trade. If there's a lot of other trading options that look very similar, then the contract is harder to default. I mean, it's harder to enforce. So in that sense, yes, it is a question of if you have these big retail platforms, how much can you really encourage people to come back and satisfy contracts because it is Amazon or a platform that has a sort of dominance over the trading space? Thank you, Jonathan. We have a few questions here from the Q&A. Jonathan has been very efficient answering some of them, but let me get one. So Christopher Cameron asked, current incumbents are markets like AIV and Compound and aren't stablecoin issuers. Is this model assuming that platforms with tie-up with stablecoin issuers or are tokens other than stablecoins considered to be the competing tokens versus CBBC? So perhaps I can quickly and then Jonathan probably gave already an answer. So essentially what we do in this paper, we don't just assume stablecoins because if you don't assume stablecoins, then the platform has more policy space. So it can also strategically decide how much new tokens to issue, what the money supply will be. So there's a broader strategy space. So far, we didn't look into that. We just never did down, okay, the money supply growth is the same as in the dollars or there's a stable relationship in terms of price. And so, but that is another way you can open up, but you open up a whole new dimension in that sense. Jonathan, did you say anything different or? Not particularly. I mean, yeah, we did start with a model with this signage benefit where you're trying to attract people to your currency. I think that we decided it was a little bit orthogonal to some of these questions about interoperability, at least to a first order. We brought it back just to stablecoins and focused on this question about, how do you want to make this system interoperable with other systems? Okay, so there's one more question. How could a CBDC even prevent interoperability? From a technical perspective, it seems pretty challenging. It's more like a remark. So prevent interoperability. I mean, CBDC essentially means that if CBDC is legal tender, then any platform is forced to accept the CBDC. So I can always easily switch and say, okay, I will sell. That's essentially what Charlie said. So I will sell my good for CBDC. And then I cannot really enforce the credit contact where the platform automatically collects and sees the token coming in and collects it right away. So you can't really enforce the contracts directly. So the CBDC is undermining this commitment power. The platform is actually providing and hence allows more credit. So that's a drawback of CBDC. Of course, it has this advantage as I said, if there's no redemption fee essentially in this sense. So it leads to less market power for the platforms or to lower markups in the retail platform. Patrick, I think you have a question. Please go ahead. Yeah, I was wondering. So I stumbled over interoperability mostly in international context. And I was wondering whether your model is applicable to the situation. Probably not the model with credit, but the model without credit. Does this provide a framework to think about interoperability between different CBDCs as well? So we haven't carefully thought about it. In general, the big difference here, this platform is behaving strategically. So it really tries to extract a rent. I'm not sure whether other countries have the same motive. But I think a certain strategic element can be there. That's true. So I have to think more about this. In theory, I think the framework could be applied. It's probably more difficult to switch to another country. So the fact that you're born in the certain token. So this entrant token is not there. That's also the case in international context. You're born in a particular country and then you automatically get your wages in this thing. So this locks you in. So the lock-in effect you have, which we have in the first part of the paper as well. You also have an international context. So we have to think a little bit more. We haven't really thought carefully about the international dimension. I think that is probably something you can extract from there. Jonathan, you have a take on it? Yeah, I would echo that. I think that the lock-in effect is particularly powerful when the currency is the default. So by default, everybody is accepting and holding this particular currency. That makes it very hard for an entrant to come in. And that is quite applicable to, say, the national currency in a country where it tends to be the default currency. So I guess in that sense, you could think about the difficulty of trying to usurp a different national currency. The initial national currency would have a strong advantage here. There's a strong incumbency advantage in being the default currency. Analysts, anyone more questions? Maybe let me just throw one here. I mean, you argue that as CBDC comes in, the smart contracts are harder to enforce. I mean, it sounds to me, ultimately, and I must confess I came a little late to this talk. So maybe I answered this previously. But ultimately, it sounds to me like a limited enforcement type issue, right? That if you can't commit to something, then I think we know from contract theory that you should make an upfront payment, right? You should post a bond, right? So the question is which side is committed? So it looks to me here that the platform is offering a service. And if people want that service because they can match better on that platform, but then a CBDC would destroy possibilities, couldn't that be resolved by having people come to the platform, make an upfront payment to the platform operator and say, now give me access to the platform and we can have all these smart contracts into force. And if I don't obey, you take away my bond that I pledged initially. Yeah, that's correct. So what we assume, as I probably didn't stress that you come in with nothing essentially, you have to buy this input and you have to take some credit because you don't have any additional resources to pledge as a collateral. So the idea essentially is the platform itself by having this better enforcement technology is itself a collateral in a broad sense. And that's what we see smart contracts essentially as a way to provide a commitment device. And you want to stick to it if you have a higher matching grade with this platform compared to the platform. So you would like to sell it because you find more easily some opposing party party you can trade with later on. And this itself gives the platform more collateral power, more commitment power, and that you would like to offer. So it's often said, you know, when you talk to people from Alibaba and all these guys, information is the new collateral. And it tries to reflect this to some extent that the platforms have this ability to offer some credit because they can enforce the credit more easily. Jonathan, do you want to? Sure. I mean, the agents want to be able to collateralize future sales revenue, like, you know, future accounts payable like a bill of exchange. The observation is that under two conditions, one, people find it more desirable to sell their goods on the platform, right? So, you know, not being able to sell on the platform is costly. And secondly, you can write a smart contract that dictates what currency you're going to receive the payment in. So that the platform, even though they're not intermediating the trade, the platform can kind of see the ledger and is able to write a smart contract and where the funds are going to go. In these situations, the agents are going to be able to collateralize more of the future sales revenue. More questions? I don't see any comments. So maybe I can ask a question about the informational aspect. You already mentioned how the platform maybe can use the information to provide credit. So sometimes people argue that this platform, maybe Alibaba or Facebook, they try to issue a payment instrument to collect more information. To enhance their, say, targeting services, right? So because when people use their payment instrument, they may use it also outside of the platform so that that allow the platform to acquire even more information to enhance their, you know, kind of to enhance data generation and monetization. I don't know if that would be something relevant to your model. Yes, so the information the platform has, it just sees the transaction paid with the ledger. So in a sense it has a better enforcement technology through this ledger payments. It sees the payments on the ledger. So we don't have this element that the platform might see even outside of the ledger a lot of activities and can combine that. So this element we don't have. We don't know that length about this to this informational advantage. But then people told us, you know, Google can see every email and they can analyze what purchase you did online as well. So the question is, does the platform have an edge over Google who can look through all your emails and all your essentially all your payments as well. So then we weren't sure whether the informational stories the key and we thought much more this commitment stories is key. So some information component that they'll see when you pay through the ledger than they'll see it. Yeah, so it's very nice framework. So maybe can I ask another question? Maybe one way, maybe the CBDC or central bank think about interoperability, maybe the CBDC providing the infrastructure for platform to provide smart contracts. In this case, the platform just specialized in competing in the smart contract dimension and then but they all feel riding on the CBDC infrastructure. So what do you think about that? Maybe there's a better game for the platform to do that. So maybe the model is useful in think about that kind of the question. Yeah, so you can you couldn't think of that. Yes, some CBDC structure is the infrastructure and summer is designed in such a way that you the blood from the private platform can still enforce the credit contract. So somehow the information and enforceability is guaranteed. So whenever something is paid on the CDC infrastructure, it is also the information is passed on to the platform and then enforced somehow enforcement is possible. So we haven't thought about this so carefully, but that's an interesting point is essentially a little bit like the Indian system where you have two layers. There's a big bone layer done more on the CDC side and then there's top layer loaded on that. So we haven't thought about this so carefully at this stage. That's something worth exploring. That's a great question. I think I think at a high level. So we have a payment system that bundles activities. So some people have pointed out that, you know, it bundles provision of deposits with provision of credit lines. And if you add CBDC that only gives the deposit side it doesn't give the credit line, you lose any potential benefits from this bundling. And here we're trying to point out a different kind of bundling that could occur, or, you know, that does occur in the smart contracting space to bundling of being able to have information about transactions on platforms with being able to write contracts on the ledger. And we're trying to play out something about the consequences of bundling that. And yeah, I think it suggests that you would want to think in your, in your regulation of CBDC you'd want to think as much about sharing information about where payments are taking place with the CBDC ledger as you would want to think about just providing CBDC itself. I think that's the insight here. I think that's a great question. And the, the, the equivalent of that story does happen at clearing and settlement at wholesale levels as well with this question of the bundling and bundling of versus the netting and the and the cheering that goes on at that level too. Yeah, thanks. Can I ask a question. I believe at the beginning. One of the authors mentioned that it was not given by senior Raj. So, if you look at the central bank balances today, very large with no time table to unwind. The moral work if table coins, backed by central bank reserves only, not by other HQ LA if table coins backed by central bank reserves are a proxy to CBDC and the technology, which largely lies with fintechs and non banks can be used. They have interoperability route, because banks don't have a vested interest, given the interneting etc. I did not follow the moral fully but I believe, could this change bridge the gap between the paper and Mr. The last two comments. Thank you. Yes, I think you're pushing the whole thing even further. We essentially go for a stable coin arrangement, how the stable coins are backed, whether they have a central bank reserves or not. We don't go into these issues or we take this as given so the people is not going so far to really look into that as well so of course you're rightly pointed out that center banks balance sheets are very large at this moment so there is a lot of space to go on the side for central banks as well to issue something which then stable coin creators can hold as a narrow banking arrangement. So we did not go into this world too much because we tried to focus here on the interoperability, but these are also important issues one has to keep in mind. Yes, so we have one more question from the Q&A. It says, but the issue of CBDC destroying commitment is more about its legal tender status than being a CBDC. What in the model differentiates CBDC versus legal tender? No, it is really the legal tender status which is the acceptability which is the key that you can just go move away and switch over to the whole thing and CBDC very easily and that the legal tender aspect of CBDC is the key. Otherwise the introduction of CBDC if it's not legal tender, then it's less of an issue. And there's also, of course as I've learned recently, there are different definitions of legal tender. There's not a single definition but it's not a continuum. One can probably work on this where we say okay, we limit the legal tenderness in a sense to such an extent that we don't undermine the credit provisioning too much. So that's something to be thought about. We haven't thought about it, we just thought total acceptability legal tender, CBDC is legal tender, total acceptability. And perhaps one can make it more continuum as well when one thinks about it. For the model I wouldn't go away from that but for the interpretation on the policy implications one can think about along these lines. So thank you very much, Marcus, Charlie, thank you very much for all the answers and clarification. So I'll turn to Todd, close. Okay, yeah, so thanks everyone thanks Marcus and Jonathan and thanks Charlie for interesting presentations on lively discussion. We'll join us again next month on Thursday, December 16. Our next session will be hosted by the Federal Reserve Bank of Richmond, and Wei Zhang also from Princeton will be presenting he'll talk about the data privacy paradox and digital demand. Until then, thanks everybody. Thanks a lot to particular chance. Thanks a lot Charlie. Thanks to everybody. It's been a good scene.