 Hello, everyone, and welcome to the May edition of the virtual seminar series on central banking and digital currencies. Before we start today's session, I want to make a brief announcement, as you can see on the screen here in November. The RICS Bank will be hosting the second conference on the economics of CBDC, organized jointly with the Bank of Canada and with a small measure of support by our virtual seminar series. I think this conference will be of particular interest to people in the audience, and I hope that if you have a paper on CBDC, you'll consider submitting it. As you can see the submission deadline is four weeks away on June 15, and for more information about how to submit and about the conference itself, the QR code here will take you to the call for papers on the RICS Bank webpage. Turning to today's session, our moderator for today is Larry Wall from the Federal Reserve Bank of Atlanta, and so now I'll turn it over to you Larry. Good morning, good afternoon, or good evening, as whichever is appropriate. Today we're delighted to have Christopher Birch, a senior economist in the research division of the Swedish Central Bank, the RICS Bank, present his paper Stable Coins, Adoption and Fragility. He will have 25 minutes. Christoph, the floor is yours. Thanks a lot Larry. Let me share my screen. I think now we should be able to see the slides. Thanks a lot Larry for the kind introduction and thanks to the CBNDC webinar organizers for having me here, and especially also to Alex to take the time to prepare a discussion on this paper. My paper is on the topic of stable coins and it's motivated by the rapid expansion of the stable coins market, the inherent fragility of stable coins, and their role as a link between crypto markets and traditional financial markets. The stable coins started to get a lot of attention after the Facebook Libra announcement in June 2019, where Facebook announced to launch a global stable coin back by a basket of currencies, and this was a wake up call for the regulatory community. While the Facebook Libra project was ultimately discontinued after the regulatory headwinds, the stable coin market continued to develop and stable coins play a critical role in crypto markets. Today's dominant stable coins are tether US dollar coin and Binance coin. They are all packed one to one to the US dollar and fully backed stable coins. The paper is motivated by the following three questions. First, what are the reasons behind the fragility of stable coins and what factors contribute to their fragility. Second, can stable coin adoption be excessive and third, how should stable coins be regulated. To address these questions, I developed a theoretical model of stable coins that allows me to analyze the determinants of stable coin adoption and fragility. For this framework, I'd like to contrast the beneficial role of stable coins with a risk of stable coins runs and potential downsides from wider adoption. Furthermore, I would like to shed light on prominent features of the stable coins market, such as the role for payment preferences, transaction costs, network effects and all hazard problems with issues. And last but not least, I'd like to gain insights for the risk assessment and appropriate regulation of stable coins. In the nutshell, what I'm doing is I'm building a two period model. There's going to be an interim date, which is featuring a stable coin conversion game, which is modeled as a global game of regime change. And the next underdate, where I model a stable coin adoptions game, which is built on the premise that stable coins offer a benefit for certain use cases to stable coin holders. And the potential stable coin holders are heterogeneous in how much they benefit from different means of payment. So the consumers when deciding whether or not to dot stable coins, they are trading off the potential benefits of stable coins with a return differential relative to deposits and the risk of devaluations of stable coins. In my model, I'm pretty agnostic about how exactly stable coins are generating a benefit for their holders. The assumption has a clear motivation for stable coin adopters who are interested in crypto applications. Their primary motive may be to economize on transaction costs in the crypto universe they might have also preferences such as such as a law for anonymity, but there might also be real world applications such as low cost remittances and currency substitution in emerging markets going forward. So this is the placement in the literature. So if you know the banking literature, then you have seen often the dynamic depict type of framework which has been used a lot. There typically the focus is on the asset side where banks are choosing the assets to trade off the returns with liquidity provision and run risk. So that is focusing more on the liability side in that I inductionize an adoption game for stable coins and deposits at the ex ante date and relative to other papers on stable coins. A distinct feature of my work is that I focus on the payment aspect and study the determinants for adoption and fragility with a view of the risk assessment and appropriate regulation of stable coins. The first side is the summary of the results. I find first that they are two mechanisms that can justify the concern in the regulatory community that there could be an excessive adoption of stable coins the first mechanism relates to an uninternalized destabilizing composition effect and the second one to an uninternalized network effect which can undermine the role of bank deposits as a mean of payment. Fragility is taking a central focus in this paper, and I find that factors that promote stable coin adoption, typically also tend to make the marginal stable coin holder less frighty which means there's going to be less stable coin runs. So factors that promote adoption tend to be enhancing stability of stable coins. Moreover factors that increase the revenues of stable coin issues are also promoting stability, as well as congestion effects which play an important role in the crypto universe. Last but not least I find support for regulatory disclosure regimes and rules on reserves and the capitalization of stable coin issues. You through the model. The model has three dates and the unit continuum of risk neutral consumers index by I, they're each endowed with $1 and they want to consume at the end of the game at day two. There's a homogeneous and divisible consumption good that's produced at day two by concept competitive consumption good sellers, which produce the good at the unit cost of $1. They are two monies insured bank deposits and stable coins. Consumers can transfer they get $0 endowments to date two by either holding insured bank deposits, or by holding stable coins by adopting stable coins. At day two, the sellers of the consumption good are accepted deposits or stable coins of equal value. Since I'm not interested in the banking side. I model deposits as an outside option, which gives an exogenous interest rate are deep and stable coins instead are issued by a monopolistic in issuer and I go into more detail. The stable coin issuer is offering to convert cash into a digital token and vice versa at a one to one conversion rate at date zero one and two. And this is not the optimal contract here, but this is what we observe in reality this one to one conversion promise where the leading stable coin if issuers all operate a unilateral exchange rate pack to the US dollar. In the model, the stable coin issuer may not be able to keep her promise for one to one conversion at all dates. And that's because the funds collected at date zero die invested in a risky technology with a date to return of theta, which is to hashtag following a uniform distribution. And importantly, theta lower bar is smaller than one meaning there are states of the world where the issuer is unable to redeem all the coins at par. The important ingredient, my model is the demand for stable coins, which is generated by an interplay of two factors first payment preferences, and second conversion costs. Specifically, consumers faith phase edus in credit risk about their consumption preference at day two, and the sellers they have a payment preference. So they might only accept one money or another money. However, they are conversion costs at each date. And importantly, I assume that the conversion cost at date zero is smaller than the future conversion costs tall one and tall two at future dates at date one at that day two. And this is generating an advantage from having the right money on hand at day two. The advantage from making your conversion at date zero, if you believe it's very likely that you're going to need a certain money in future because in future, it's going to be more expensive to do the conversion. Here it comes in as follows. They are capital she groups index by small G, where capital she can be arbitrarily large. And an individual consumer belonging to group G has a probability alpha G of being matched with a consumption good seller who only accepts payment and stable coins of probability better G to be matched with a seller who only accepts bank deposits and a probability alpha minus probability one minus alpha minus beta to meet a seller that accepts both types of payment. Importantly, if a consumer belongs to a group with a higher G, then she has a higher expected need for stable coins because she thinks it's more likely she meets a seller that wants to be paid in stable coins. So the consumers belonging to groups with a high G are going to be the crypto enthusiasts who are most eager to benefit from stable coins, whereas the ones with a low G are more inclined to hold bank deposits. At date zero, there is the adoption game as I said previously and that's modeled as a simultaneous decision where consumers to decide whether to adopt stable coins or instead to adopt to hold insured bank deposits. At date one, there is this conversion game modeled as a global game of regime change. And importantly, if the issue is meeting redemption requests, liquidation is costly. So she has to liquidate part of her long term investment and the premature divestment yields a return of little R, which is smaller than theta lower bar. So meeting early redemption requests is draining the resources of the issuer. And if the issuer is unable to meet her payment obligations, there's an additional bankruptcy cost psi. To close the description of the model I would like to point out an important assumption that has been used also in the literature as a modeling trick. I assume that coin holders are active only with a probability copper, which is smaller than R. So, in this way, I exclude rationing at date one so at that one the issuer is always going to be able to meet the redemption requests of active coin holders. But of course, at day two, the issuer might not be able to meet her promises and might go bankrupt. So the model can be relaxed, but the model becomes much less tractable, but key insights by large go through. Let me skip this graphic illustration of the model and come to the analysis. So the game is soft backwards first we look at the conversion game at date one, and then we look at the adoption game at date zero. How does the conversion game look like. So when solving the conversion game, we take the adoption rate, which is denoted as capital and as given. And we can find using the global games methodology that there exists a unique threshold equilibrium. That looks as follows the issuer faces a run at date one for all fundamentalizations theta that below that fall below a certain threshold theta star, which is given implicitly by this condition here. And this condition here is essentially an aggregation of the group specific indifference conditions of the stable coin holders. Also, why you can see here that this term here gamma upper bar is reflecting the weighted average over the payment type of the individual coin holders. So this is basically the weighted average of the gammas over the coin holders. And for that reason, the composition of the stable coin adopters is going to matter for stability because we can relate this. Gamma upper bar to theta star which is governing the probability of stable and runs. Let me come to the key comparative statics, which you can see here in this table. The first three comparative static results are consistent with what you would expect from the banking literature which gives us some confidence that the stable coins run model is plausible. First, I find that an increase in bankruptcy costs associated with a higher probability of bank runs, because there's less resources available in case of a run. So coin holders are going to become more flighty. The fraction of active coin holders copper is also associated with a higher probability of stable coin runs, because they exert more pressure on the issuer. And then an increase the liquidation value instead is stabilizing it's reducing the probability of runs because a higher liquidation value gives the issuer means more resources to stem against redemption requests. The fourth comparative static is one that I think is relevant for the crypto universe. The probability of runs is decreasing in the conversion cost that they do on. So if for instance due to congestion effects. The conversion cost that that one goes up and I indoctrinize it in an extension also of this model for the time being it's exogenous. Then we have that the probability of runs go down so that's a stabilizing effect here in the system. Last but not least, we have a comparative static result on the composition effect. Specifically, I find that the average relative preference for stable coin payments is negatively associated with the probability of bank runs. Why is that the case. This is reducing the probability of bank runs. Why is that the case. If I have stable coin holders that are more eager to hold stable coins that have a higher gamma. Then this is driving up the average gamma in the pool of stable coin holders. And as a result of that, it's reducing the flightiness of coin holders in the conversion game and reducing the probability of stable coins. And this is a destabilizing composition effect that's going to be also at the core of efficiency result which I will show you in two slides. After having looked at the conversion game at date one we can go to the stable coin adoption game at date zero. In date zero when deciding whether or not to adopt stable coins, the consumers, they are looking at their group specific benefit from adopting stable coins which is given by this formula here. They essentially look at the expected return when adopting stable coins and compare that with the expected return when holding bank deposits. And interestingly this, importantly this benefit from stable coin adoption is increasing in the groups. So if you're more likely to meet a seller that only accept stable coins of course you have a higher benefit from holding stable coins. It's decreasing the deposit rate, higher deposit rate instead makes deposits more attractive relative to stable coins. And the benefit from stable coins is also decreasing in theta star so it's decreasing in the probability of stable coins. So if there's a higher risk of devaluation, you're also less inclined towards stable coins. And this gives me a relation between the adoption rate and the probability of stable coin runs. So we find that belief about a higher probability of stable coins at date zero is associated with a lower adoption rate. And based on this destabilizing composition effect from the previous slide and this comparative static, we can do an equilibrium analysis, and we can find that there exists a unique equilibrium of the adoption game. Now, based on this results as a next step, we look at the efficiency analysis. And here, I like to address the concern of regulators about a widespread adoption, perhaps rapid adoption of stable coins. I think about Facebook Libra for instance. And I want to look at that through the lens of my model and try to understand whether there could be scope for inefficient excessive adoption. And there I focus on two internalized effects first the destabilizing composition effect that we just discussed and second, an internalized erosion of bank deposits. So coming to the destabilizing composition effect, let's have a look graphically at a model where there are two groups of agents, group one and two. And what I do here in this graph, I'm plotting their respective benefits or the payoff differential from adopting stable coins against the adoption rate. What you see here is that the benefit from adopting stable coins starts sloping downwards whenever multiple groups of coin holders are adopting stable coins. And that's because of the destabilizing composition effect which makes it less attractive to adopt stable coins in equilibrium. It's going to be the case that the coin holders here. The marginal coin holder belonging to group one here is going to be just indifferent between adopting stable coins and not, and we cannot compare this equilibrium adoption rate to the constraint plan of problem. And we find that the marginal adopt of stable coins is not internalizing that she poses a negative externality on other coin holders through the destabilizing composition effect. Graphically, you can see that the benefit from adopting stable coins in green is rather small compared to the cost imposed on the other stable coin holders by making the stable coin less stable through the increased adoption. So to sum up, we have the potential for excessive adoption due to a destabilizing composition effect that's not internalized by the marginal stable coin adopter. Now, if we extend the model and introduce network effects, this result can potentially be overturned. And the way I do this is I assume that the probability, the common component of the probability of meeting a seller that only accepts stable coins is increasing in the stable coin adoption rate as stable coins are more prevalent. Instead, the probability of meeting a seller that only accepts bank deposits is declining in the stable coin adoption rate. And if these two effects are sufficiently strong, they can overturn the destabilizing composition effect, and it can be the case that a higher adoption is actually stabilizing. So it's really important to think about both effects when trying to evaluate whether a higher adoption is stabilizing or not. There are some caveats to this when we introduce positive network effects, multiple equilibria of the adoption game can emerge from a viewpoint of regulators. That's not very desirable because then you have potential for rapid shifts in adoption because the beliefs about stability change. And we can also see that the origin of the positive network effects matters. And to dry form this message, we can develop a second reason why stable coin adoption might be too high. And that's because of the uninternalized erosion of bank deposits. For that suppose that a wider adoption of stable coins is reducing the probability that deposits are accepted. Formally, we assume that alpha prime is positive and beta prime is zero to isolate this effect. And here's the case that the marginal adopter doesn't internalize that she poses a negative externality, not another stable coin holders. But this time on the depositors on the ones who decided not to hold stable coins. Because by her decision to adopt stable coins, she makes it less likely that this depositors are going to meet a consumption put seller that is accepting the deposits as a payment. I have a few more extensions of the model to try to analyze the different determinants of stable coin adoption and stability over in the buffer what we have discussed so far. From a regulatory viewpoint, there is an interest in thinking about more hazard problems of stable coin issues, because we see in practice that some of them have rather risky investments. And there's also concerns about the quality of disclosure. And let me look at this extension. And you can refer to the paper if you have interest in some of the other extensions that you can see here on this list. So the way I introduce the more hazard problem is I introduce a portfolio choice problem where on the asset side the issue can choose between the low risk and the high risk portfolio. And the high risk portfolio choice is essentially modeled as the mean preserving spread in the distribution of fetus in combination with the reduction of the liquidation value are. So here it's clearly socially optimal. If the issue chooses the low risk portfolio, which has a lower spread in the theta distribution and the higher liquidation value and therefore also I guess rise to a more stable stable coin. However, the incentives of the issuer are misaligned. And one question is whether a regulatory disclosure regime can help to overcome this problem and to implement the efficient risk choice. I find that this is not always the case. And that's because the issuer is only implementing the low risk choice if the sensitivity of the thresholds theta star and and star is sufficiently high. This is a little bit different to what you typically know from the banking literature where with transparency the price of debt is risk adjusted. Here we have the unilateral exchange rate pack. So what matters is only is the sensitivity of the thresholds to the risk choice of the portfolio. And given that the regulatory disclosure regime doesn't necessarily help always. There are other implications for additional regulation, such as capital requirements and the regulation of research of issuers. I have another extension also on stable coins landing that's also related to one of Alex's papers, where I introduce a stable coins landing game in between the adoption stage and the conversion stage. In a way, as it's done in the traditional currency attack literature in the paper by Corsetti. And I introduced large borrow that is borrowing stable coins, which is got three minutes. Yeah, which is actually great if the larger borrower is not a speculator, because by offering some interest on stable coins. I'm not really promoting stable coin adoption because it becomes more attractive to adopt stable coins if you can earn interest as long as the stable coins are still relatively stable. If it's of course likely that this large borrower is a speculator, then this benefits might be eroded to sum up. I will post a framework in this paper that is modifying existing theories of bank grants and currency attacks by modeling stable coin adoption and by incorporating features that I think are important in the stable coins market, especially this payments role. A key ingredient in my model is the demand for stable coins, which is generated by a heterogeneity in the induced payment preference of the potential adopters of stable coins of the consumers. The main results are first, I identify potential downsides from wider adoption, specifically I isolated to externalities that can lead to excessive adoption of stable coins. Second, I provide some insights for the risk assessment of stable coins from the study of the determinants of stable coin adoption and fragility. Third, I find support for regulatory disclosure regime and roots on reserves and the capitalization of stable coin issues. And last but not least, what I didn't have time to show you here, where I have to refer to the paper, I offer a set of new testable applications. Thanks a lot for your attention and I'm really looking forward to Alex's discussion of the paper. Before we go there. There's one clarification question along treated as just clarification now if we want to go into more depth we can do that in the Q&A. Yeah, general ask the question, can stable coins choose to put their money in bank deposits as well instead of investing in a risky project or asset. A stable coin issues or stable coin holders. I interpret that a stable coin issuers. So, so, so, okay, so it's going to be the case that so that's the slide I didn't show you. So it's going to be the case in equilibrium that we're going to have a threshold on gamma. And the consumers belonging to a group with a high gamma find stable coins more attractive they have a higher probability of needing of benefiting from stable coins as payments at day two, whereas the ones with a low gamma they go for bank deposits. The stable coin issuer collects the funds from the groups with a high gamma and invest them in a risky investment project. Then in the extension when I look at the mall has a problem. I allow for risk choice, a risk management problem here of the issuer, so we can decide whether to invest in a more or less risky portfolio. And you could also model that as a portfolio choice problem where you have a risky asset and a safe asset, and you decide on the mix between the risky and the safe asset. Okay, okay, I think that. So, so in theory it would be possible for them to go riskless and put it all and ensure deposits in a too big to fail bank, say. Not my focus but in principle if you introduce the risk management problem, they can go fully safe. Okay, and I have an extension also in the paper where discuss them. E money and narrow banking with which goes in the direction as well. Okay, I think we need to move on to Alex now. So, discussing the paper will be Alexander Locas, the chief of the macro potential policy analysis section of the financial stability division of the Federal Reserve Board. And Alex four is yours. Thank you. Thank you. Thank you. Let me serve my slides so. Okay, all good. Can you see them great. Thanks a lot for the organizers for inviting me to discuss this very nice paper. So I enjoyed a lot reading it but let me jump directly to it. Starting with some basics for stable coins. What are they I mean, they're digital assets that they promise to maintain the constant price of $1 and billy Dima bill at parent demand. Now, most of the stable coins that we have out there like death there or USDC die another stable coin. They're mostly collateralized by other assets so they get their safety from them. The issue is that a lot of these assets can be risk and illiquid. Now, this is not something new in financial intermediation. This is very similar to what we would have in banks with uninsured deposit and liquid assets or money market funds. So it is very, very reasonable to expect that these stable coins will be exposed to run risk. What is the big difference of stable coins from other money market instruments like deposit money market funds? The big difference is that they don't provide any direct compensation to holders for this run risk. And what is the reason for that? Stable coins pay no interest. Now, we can argue whether this is an optimal outcome at the contracting problem, whether this is regulation because you're trying to avoid being classified as securities without external or it is an asymmetric information thing. But I mean, let's take it as given they don't pay interest and this is a very important thing. Why is that? Because then we really need to ask ourselves, OK, if they don't pay interest and they are risky, where does the demand for stable coins come from? This is a first order question. Now, another question which is related but distinct to that is how can you maintain your peg in secondary markets trade continuously? So this question, of course, are related, but they are a little bit distinct. This paper tactics very nicely the first question and focuses on the role of stable coins to facilitate payments. And I will get more into that because I think there are many novel aspects in this thing. But at the very high level what is the paper saying? The paper is saying that if stable coins can be used more efficiently for some time of payments than other private money, then there will be demand for them. You will demand these things. And how can we actually, this is a quote from other people. So the idea is the following. If I am only willing to exchange my dollars in a foreign country for the local currency that has the risk of devaluation, we can imagine these currencies. If I know that I will need local currency to go and buy a coffee at the local store, otherwise I will not do that. I will use my credit card. So I will only be willing to take the risk of devaluation, inflation from one day to another, or you can think of this, if there is demand. And that's where the paper starts from. Now, Christoph already mentioned this in another paper with Gary Gordon and people from the Fed. We are investigating a complementary channel where the demand for stable coins comes. It will be to take leverage speculative positions in crypto empirically and theoretically, but I'm here to talk about Christoph's paper, so just a little bit there. So what is the sketch of the model? So there are three periods. There is a simple tractably model. There are three periods. There is a stable coin issuer and there are heterogeneous agents with respect to their payment preferences that they choose between insured deposits and stable coins. What happens in this period? At equal zero, the issuer caters the demand for stable coins and invests the process in a single risk and the liquid project. There was already a question about that. Christoph relaxes this in an extension and considers it a portfolio problem. As an aside, I think this is the first of the risks that you should elevate in the paper, so to solve the total optimization problem. I think it doesn't belong in the excess, it belongs to the main part of the paper and you have already done the work. Now, at equal one, some agents become active and some become passive. The active agents decide whether to redeem their stable coins in a global game. This is a typical demo, difficult, standard stuff. There is nothing weird there. They borrow a trick from actually the mutual fund paper that Christoph said that say that only some people are active. Now, I never like this assumption but people are doing it. In this particular case, I think it's a little bit problematic if we want to be realistic how to think about trans stable coins. And the reason is that these things are trading 24-7 decentralized blockchains. There is full transparency about the price. You can even set up and alerting your smartphone app to see where the price goes down. And there is a lot of social media presses. Now, I think this is not to criticize your paper. In general, I think it's a very important aspect to think that runs happen fast. And this is what the March 2023 episode saw that's even for banks. Bank runs have much faster than before. So we shouldn't forget about that. Now, T equal to if the stable coin is solvent, agents can use tokens for certain payments with some probability. If it is insolvent, they get the process. That's the model. Now, the key thing of the paper and I think the big novelty of it is the payment type probabilities. So what's the key aspect? There are heterogeneous preferences such that with some probability, they prefer goods that require payment, either stable coins or deposits. There is a common component and the idiosyncratic component is probabilities. Let me start with the syncratic component. The idiosyncratic component ranks each agent individually such that some people have higher probability, some have lower probability. It's completely exogenous. The first inefficiency that Christoph shows comes directly from this idiosyncratic component. And the idea there is the average pool will determine their own probability. But as you add more people that they have lowered these probabilities, you deteriorate the pool. The probability goes up. This is an externality. This is a novel externality. It's very nice. I like it a lot. There is another common component, which is what we call the network effects that is increasing on the stable coin circulation at equal to. This gives you the d an alpha n or the beta one minus n in the slides of Christoph. Now, I'm less enthusiastic about this component for two reasons. First is a pure modeling theory thing. It is inconsistent with the number of coins circulation in not out of equilibrium parts in the global game. So this component only depends on the number of circulation at equal to zero. If you want to do it properly as a common component, you have to introduce it in the out of equilibrium parts of the global game. It is doable, but it is worth. The more important reason why I'm less enthusiastic about that is it's a little bit of hope. And I think it is very when we talk about contagion externalities. It is a contention in the old search version of the one that you saw. It's very important. You saw in the blockchain, you know, with the gas fees. It's very important to micro found it. And there are two ways you can do it. You can do there. You can do random search with bargaining. You can do directed search either way works. I think it's very important because we need to know which way the matching efficiency goes and which way the congestion externality goes. So I think this is an important part if you want. I mean, it's an important part. It's really there. Now it's also doable because you can do network effects by increasing returns to scale matching functions. There is a paper by Antonio Coppola. Chris Namurthi and Jackson suit that they do this. So you can have a look at this. It's actually a currency adoption paper as well. Not the stable conduction currency adoption. So they do it with increasing the scale. I'll exceed that three minutes. Fantastic. Thank you Larry. Now, as far as I'm concerned, there is already a lot of novelty in the first efficiency you saw. And I personally would be happy to leave it there, but you know, the network effects do not seem to matter for keen sites. So you may consider, you know, pushing it as in the end because there are some. There are these issues. I don't know if this may be me or others, but you know, there's this issue. Now, the second thing that actually, you know, it gets just running. I mean, it's a very nice paper. So you want to say, OK, this is the adams. You understand this happening. OK, the theoretical point is very straightforward. We get it. But the paper may benefit a bit from the quantification of the mechanism to see how economical significance. And that will get you. I'll explain why. Let's go to see the main premise of the paper is that table coins are used for payments. So there's a convenience here. Right. This is what gives them their demand. How is how big is this convenience here? If we go to a paper by Scander on the convenience of deposit, it is about 80 basis points in recent years. So if we think that this is a very reasonable estimate or an upper bound, if you like, for the convenience for stable coins, because they cannot use for everything. Can we justify this huge increase from 20 billion to 200 billion in just one year? So I think this is the first order question as well. I'm not sure that with the underlying big risk, you can justify it. It's definitely a component. And we definitely component if Libra is adopted. But here for this market and the policy becomes for this market. Anyway, it's important to quantify. So let me say overall, this is one of the early papers on global game approach to stable coins. It's a very nice paper. I would really recommend that you go and read it in detail. The focus is on the role of stable coins for payments. And I think you're making a very nice contribution with the heterogeneity on the payment types. Thank you so much. You have brief comments and response. Yeah, thanks a lot for this great discussion. So that's definitely food for thought. Maybe a few thoughts. So regarding the convenience yield, I also think it would be really important to try to learn more about that. It's an empirical question. And from my reading of the literature, generally, the use of stable coins and crypto has a lot to do with illicit, illicit activities, money laundering, organized crime and so on. So I guess that can be very strong reasons that lead to adoption for some, right? And that can also potentially explain rapid growth. But it's very difficult to say and I don't have the insights empirically. But that's definitely something interesting to look at. And from the viewpoint of my paper, I really tried to be a bit agnostic about the different use cases. There will be a very strong of different groups of consumers with different use cases and maybe going forward, there will be a very strong use case in currency substitution and emerging markets if providers are penetrating that market and that could lead to a lot of adoption perhaps or not. I also think that your comment on the network effects is really great. And that's definitely something that I have to think a little bit more about and have to see as you said how I positioned in the paper and how much deep I want to go into the micro foundations. Admittedly now it's relatively ad hoc and it's modeled a little bit as in the paper I cited by Itai Agua and co-authors. And there's definitely scope to do more on that if I choose to go in that direction. Maybe it's good to get some more bilateral feedback from you to see whether it's worth to do it. Thanks a lot. Well, while we're waiting for some more questions to roll in, I wanted to pick up on one of the points that Alex raised about how stable coins having managed to maintain their peg with secondary trading. Because you model this as a mutual fund where redemption is required but in fact what we frequently see is that they choose not to redeem it, not to impose those liquidation costs on their holders. But instead, I'll let those who want to run away do sort of discount on the market. Have you thought much about that and what that might mean to the model. No, I didn't think too much about it. I mean, if you have introduced trading and prices that are revealing information in this type of model, you have to be very careful how you do it with the timing in order to keep the global games framework going and get a unique equilibrium of the conversion game. That's that's one aspect to keep in mind but there are possible ways to do that. It is the case that the leading stable coin operators, they have certain agents that are helping them to do this open market interventions in order to stabilize the pack and there are certain types of arrangements. And that's also something that I would have to learn more about first to better see how I could potentially address that. Francesca, you have a question. Hi, yes. Thank you very much for the presentation. I have kind of like two questions. One is more model specific and one I guess it's broader. So I'll ask quickly both. The first one is how dependent is your externality effect against the network one on the assumption of a monopolistic issuer of stable coins. And the reason why I ask is that it seems intuitive to me to think that there may be an incentive for stable coin adopters or crypto enthusiasts, whatever to sort of partition themselves into different sets into different groups that uses one stable coin versus another one. So for example, the, the more enthusiasts or the ones who value their privacy the most, or the ones who are sympathetic with Russia the most. They might be more willing to use tether other ones might prefer to use us DC. So that externality. I wonder how much that externality effect depends on your modeling choice of having one stable coin and one monopolistic stable coin issuer that doesn't, as far as understood, that does not allow this sort of like partitioning different sets of stable coins or monies adopted in equilibrium that might have different characteristics about this externality effect. And the second one I guess it's maybe a bit unfair but and I'm sorry if I missed that, but what is really in this model, the element that distinguishes stable points why is this not a model of banks, instead of stable points versus deposit why is this not a model of bank deposits versus cash. Let me maybe start with the first question so. So that's really interesting so I have formally looked at free entry to the stable coins market but still only at a model where where you know you have one stable coin issuer, but free entry. I, I have been thinking also about multiple stable coins. Then, in order to get the type of situation that you describe where some might prefer one over the other. You would have to introduce some additional element in the model. Maybe, you know, move away from risk neutrality or something in order to have some preferring the risky gambling stable coin. Whereas others go more for the safe stable coin right, but that's definitely something that's potentially relevant and interesting coming more to the core of your question regarding the role of the network effects and all that. If you believe that the network effect has a market specific component and a issue specific component, it would be probably important to to think which one is more important than the other. What I did in the paper in an extension I looked at the fixed cost of operation, which you can, to some extent qualitatively interpret in similar ways so if you are large stable coin issuer, you are better able to cover your fixed cost of operation than a smaller one right, so so so that gives you also some gains, but that's something that's definitely worth to discover more and for some of your questions I think we need additional ingredients in the model to get at it. Your other question was about now I forgot what it was about. It was just, it was just about what is it really in the model about stable coins why is this not a model of instead of stable coins versus deposits of banks and bank deposits versus cash. Yeah, that's that's also good question so so generally, when I started to think about this this this model I was really intrigued by the observation that the crypto crowd seems to be very heterogeneous. And that's why this idea from from the construction of the demand side comes from to the extent that some something like that also matters for traditional banks. And it also applies in a way there I mean if you think about literatures on institutional investors and so on that might be also different groups of investors that have different benefits from certain investment products and that that might be more or less flighty. So to that extent it could also apply more broadly, but I didn't think about that carefully and it was really the motivation that this crypto crowd is really heterogeneous that that was guiding me and how to model that. So, I see a pretty facade hands are, this or her hand up. Did you have a question that you wanted to ask your unmuted. I think you can unmute yourself. Okay, either I'm not handling the technology writer there's not a question here. Let me switch back to the to Todd to ask a question. Okay. Thanks Larry. Yeah, Chris, I think this is super interesting I wanted to follow up kind of on on Francesca's last question. So, you know, like, thinking about stable coin adoption. So what is the comparison or what is the relevant decision margin. So, so you've kind of got a model where everyone's thinking about, I'm either going to use stable coins or bank deposits for something and then different people would value those differently. I have to admit I sort of struggle with who uses stable coins and why but you know if I think that at least part of it is about, you know, I want to trade on the blockchain. I'm speculating on Bitcoin versus something else and stable coins are sort of a useful way of, of running those trades. And for those types of people or bank deposits irrelevant comparison or, or, you know, you know, because the bank deposit you can't, you can't trade on chain. For them is it more about if there were no stable coins I'd just be trading Bitcoin versus ether, instead of Bitcoin versus tether and tether versus ether. I wonder what are your thoughts on that how important is the stable coins versus bank deposit deposits margin or how would we get a sense of how important that margin is versus stable coins versus some other crypto asset. That's a good question maybe Alex is actually in a better position to to answer it given that his paper is making a link between the speculative demanding crypto and the demand for stable coins right, but I think that that generally, if you think about the speculative speculators retail speculators, I mean, one casino is to to to go into, you know, this game shop type of stocks right or big tech stocks and another casino is a crypto and for one casino you use fiat for the other you have the stable coins so so there's certainly some users that, you know, use both. Generally, as you said, there are specific reasons why a lot of crypto investors use stable coins because they can reduce their trading costs in the crypto universe. And my model is in a way stylized but you know when I talk about a seller that is only accepting stable coins and your consumption preference to buy from the seller. So the idea of also mean to encompass is a situation where you know you are likely to be interested in in doing some speculative crypto investments in the future you know that already now. And that's why you you're adopting the stable coin right. So, so I try to interpret that relatively brought with it broadly. I have one question Larry to Crystal. I cannot raise my hand as a whole so I'm wondering. So like Alex last also, it made me wonder it may be hard to justify with the 80 basics point as a convenience user I'm wondering maybe one motivation is about the heterogeneous believe. The crypto are super optimistic about crypto or stable corn that driving on top of the 80% 80 basic point. They may be really cool for it so but is it any different with the heterogeneous believe and heterogeneous demand in your model. Are they the same thing or have some difference. I understood your question right now there's no link so so I mean stable coins don't really offer an upside. As long as we we don't introduce a stable coin lending so that you can earn some interest right. And yeah, so, so there's no rule for you know optimistic optimistic beliefs on the return of stable coins that are justifying stable coin adoption. Some people are worried about the downside of crypto so they want you to park their money in something stable that they provide a safe haven that kind of heterogeneous believe on crypto. That may drive people to hold stable corn I'm wondering maybe it is isomorphic to some heterogeneous demand in your model that okay we can speak to Alex question. So some of the elements look like can justify the basic point in the data. I would have to think about that more carefully, I think it relates to thoughts previous question because then the adoption game would be basically or the choice would be x until about, you know, crypto assets that that are not stable coins like Bitcoin for instance and versus stable coin right Bitcoin versus stable coin would be an index under stage right. And that's part of what I had in mind when and we think about it that way, you know the convenience you could be more than 80 basis points right it's, it's not tether versus treasuries, treasuries you just can't use for this. So, I see Nathan Palmer has his hand up. Nathan would you like to unmute yourself and ask a question. Yes, can you hear me. Yes. A thought that occurred to me recently, listening to some podcast crypto folks talking about crypto things and there's a fair number of like CEOs of crypto companies and one of the things that a couple of them like were complaining about with those when SBB and signature were having their troubles was that they basically paid their employees and stable coin. And there was concerns that their employees, you know, us DC I think you know like DPEG for a little bit and so there is concern that their employees are going to have interruptions of their payments. I had originally just assumed that these sorts of like, I did not even think about the possibility that software engineers are being paid and stable coins. Do we know, do we have any data on, I feel like the answer is no but do we have any data on how many people doing software development type jobs or things like that are being paid and stable coin versus regular fiat. That's my question. I also heard about that before about instances where where some are paid in stable coins. But maybe somebody else in the audience knows more about that. So we are almost at noon, but we can go a little bit longer. Jonathan, you have a question. Just want to provide a little bit of information about the usage of stable coin. So we have some studies at the Canada. So according to that basically the usage of stable coin for for payment of good services very, very much very limited right now. It's only around maybe 0.043% compared to visa transactions. But the main use case is really right now seems to be as a vehicle currency for non US based centralized exchanges as a vehicle currencies and other usage is really for some margin and settlement assets for crypto derivatives. And finally, would be this stable coin is used for for as the funding asset for lending platform in DeFi. So, so given that maybe one way maybe is for us to do we think is how can we think about the role of stable coin and how in financial trade and how that may affect the measure of convenience you. Thanks. Well, I think that's pretty much it unless somebody's got another question they'd like to ask. If there are no more questions. I'll turn it back over. I guess it was. I'll take it to Todd. Okay. Thanks everyone. Thanks Christoph for a great presentation and Alex for a very informative discussion. Thanks Larry for moderating thanks everyone for attending. I'll just throw up one last time the one about the November conference at the ricks bank and I hope you'll consider submitting and please stay posted for announcements about future sessions. Okay, thanks everybody. Thanks everybody for the questions and feedback. Thanks a lot. I appreciate it.