26th Annual Financial Markets Conference Transcript: Policy Session #1: Which CBDC, if Any, Is Right for the United States? - May 9, 2022

Banner for 2022 Financial Markets Conference

Policy Session #1: Which CBDC, if Any, Is Right for the United States?

Central bank digital currency (CBDC) has evolved from a topic that central banks were studying to something that many are moving to develop and implement. Moderator Nellie Liang and panelists Charles Kahn, Paul Kupiec, and David Mills discuss how different central banks are moving to adopt the version best-suited to its mix of goals and consider whether the United States should adopt a CBDC and, if so, in what form.

Transcript

Nellie Liang: Thank you; welcome. I want to say thank you to the Atlanta Fed for the invitation to moderate this opening panel for this conference; it's great to be here. As moderator, I was given the opportunity to provide a broad framing for our discussion of CBDC and I decided to take it. I'm going to just take a few minutes here to get us set before we turn to our speakers.

I'm going to start with digital assets, which has become a topic of great interest to policymakers in the US and abroad. We define digital assets as a financial instrument that can be transferred on a distributed ledger, sometimes with the blockchain technology, to provide recordkeeping. This system enables payments and other financial services to occur, possibly without intermediaries, and that's the novelty of it.

In March, President Biden issued an executive order calling for a coordinated and comprehensive government approach to digital assets. With this executive order, agencies across the government, including Treasury, State, Commerce, and coordinating with the Federal Reserve, will analyze how to ensure the responsible development of digital assets, recognizing the potential benefits of innovation but also its risks.

Today's panel will focus on digital assets for money and payments, and in particular a central bank digital currency, or a CBDC. We will leave aside cryptocurrencies that have highly volatile prices, which are encompassed in the executive order, but are less likely to be useful for payments on a widespread basis.

To provide context for this discussion, I thought it would be useful to look at some current money and payment systems. This is a simplified table, but it's too detailed anyway, just to give you a sense. You can see on the first two rows of this very simplified table [that] today's payments systems include a blend of public and private provision of services. In broad terms, consumers and businesses use paper currency, which is a bearer instrument issued by the central bank, and they use bank deposits that are transferred by a range of technologies. Banks are usually involved in the distribution or clearing of payments, and final settlement often ultimately takes place on the books of the central bank.

Money and payments have changed over time because of innovations to address market failures, or to provide a public good. For example, in the 1800s privately issued bank notes often failed to trade at par or were subject to runs and led to market financial instability. This is a market failure. Government stepped in and provided the public good of a uniform currency.

Currently, new technology, which will be offered by FedNow, aims to offer real-time, more efficient payments to depository institutions. It does not create new monetary instruments. It will rely on the existing banking system and use deposit money.

In contrast, the next row, stablecoin is a digital asset pegged mainly in practice today to the US dollar. It could be used for payments on a widespread basis, but it resurfaces some of the old risks and market failures. They are bearer instruments, but they're issued by nonbanks. They have the potential to generate destabilizing runs if the value of the assets backing the stablecoin decline abruptly. They may also introduce novel payments system risks related to the distributed ledger technology.

So, the President's Working Group, or the PWG, issued a report with the OCC [Office of the Comptroller of the Currency] and FDIC [Federal Deposit Insurance Corporation] last year to highlight the potential risks of stablecoins—run risk, payment risk, and the potential for increased market power—and recommended Congress pass legislation to address them. Against this backdrop, we can look at "What does CBDC offer?" which is the last line in this table. What's the market failure that the public good of CBDC could address?

The next slide highlights some key values that are expressed in the executive order which are relevant to the future of money and payments. Global financial leadership to reflect our democratic values and national security are prominent. Others include economic growth and financial stability, technological advances, especially privacy, financial inclusion because of lack of access to financial services or high costs, and payments system efficiency. As you can see, there are many public goods on this list, which suggests greater efficiency of the payments system may not alone be sufficient to justify a CBDC.

The executive order also calls on the Fed to separately continue to research and report on a potential CBDC and its design features, and to consider implications for the efficiency of payments and the ability to conduct monetary policy as a macro stabilization tool. This work will build on the recently issued Federal Reserve paper on money payments, which seeks public comments on the benefits and risks of issuing a CBDC.

We have a real substantive set of issues to address today. We will start with professor Charles Kahn, who has written a paper for this conference, to provide insights from the experiences of other countries with CBDC. Then, David Mills from the Federal Reserve Board will discuss the Fed paper and considerations for the central bank in issuing a CBDC. Paul Kupiec from AEI [American Enterprise Institute] will offer a specific proposal and tell us why Treasury doesn't need to write this future of money and payments paper.

Then we'll open it for general discussion and take questions from the audience. I understand I can read them on this iPad, so I will be monitoring for your questions. Okay, with that, I'm going to turn it over to Professor Kahn. Thank you.

Charles Kahn: Thank you, Nellie, and thank you to the organizers as well for inviting me to be part of this extremely exciting panel. Digital payments are a hot topic in the US and for US regulators right now. There's been a flurry of activity in terms of reports being written. Here's a list of four important pronouncements or requests for more information that have come out in the last few months. One of them, the Boston Fed/MIT Project Hamilton report, is reporting on the technical feasibility and possible designs for a CBDC, and the others are policy pieces. I'm sure that the rest of the panel is going to be giving you much more detail about them.

Despite that flurry of activity, the US comes fairly late to the game. Here's a graph of the number of different jurisdictions around the world that have been looking at central bank digital currency over the last nearly a decade. This list is coming from estimates made by John Kiff, and you can see that we're up to, by his count, something like 80 different monetary jurisdictions looking at the stuff right now. Now, looking is not a very deep notion, so he also puts together a more emphatic version of the story, that is, which countries are out there that are actually actively doing R&D at the moment, or actively working on pilot projects, or actively actually having their own CBDC out there at the moment. This is his list; other people do it different ways.

The Bank for International Settlements also gives you a nice picture of all the countries that are busily involved. The proposals take many forms, and they advance a variety of policy goals, and because of that it can be very confusing to try to understand what's going on. I thought it would be helpful in the time I have to examine some concrete examples of countries that have done quite a bit already at various phases of the story, and to use those examples to see what messages they might have for the United States. What I'm going to do, briefly, is tell you a little bit about what's been going on in four different countries, Sweden, Canada, the Bahamas of all places, and China, and see what we can learn from those experiences.

Now, if we're going to do that, like any good academic I've got to start with definitions, so here are my definitions. Actually, the definitions that come out of the BIS. A central bank digital currency is a digital payment instrument denominated in the national unit of account and that's a direct liability of the central bank. That means that, for example, cash would fit all of those things except for the fact, of course, that it's not digital, it's physical. It differs from, for example, commercial bank money, that is, checking accounts, which are liabilities of the commercial bank and only directly backed by reserves of the federal government.

It differs also from stablecoins, which are digital assets provided privately outside of the banking system. While they're designed to maintain a stable value relative to the national unit of account, they're also not direct liabilities of the central bank.

What about central bank reserve accounts? That seems like an odd account. They're digital. They're fixed in liabilities of the central bank. They're used by commercial banks to pay each other. Seems like that ought to count right down the line, and we can just close up and go home and say, "Yes, we've got a central bank digital currency already. Let's go out and have drinks now."

No. Very helpfully, various authorities have explicitly left that outside and said, "We're talking about something that's not a traditional reserve or a settlement account. We want something a little fancier than that as part of the story." If we didn't put that extra caveat on the story, then reserves of a central bank would very closely match one of the basic kinds of central bank digital currencies, which is wholesale CBDC. Wholesale CBDC is the kind of central bank digital currency designed specifically to facilitate payments by financial institutions of one sort or another.

Canada has examined wholesale alternatives like this very closely, but most of the other countries that have been examining CBDC have instead been looking at retail or general purpose CBDCs, that is, assets intended to be used not for payments between financial institutions but payments by the public on a day-to-day basis.

That's the first grand division of the kinds of CBDCs that are out there. Once you go beyond that, there's a whole pile of alternatives that are proposed or considered. First of all, there are CBDC proposals out there that include interest-bearing assets, and others that say, "No, we want our asset not to be interest-bearing." There are direct or delegated versions of this story, where direct means that everybody in the world would have an account with the Fed, or everybody in the United States, at least, would have an account with the Fed. Or maybe instead we should all be having money from the Fed, but it's intermediated by some institution that provides the services that stand between us, the public, and the Fed itself.

They vary in terms of the guts of the workings. Some of them propose centralized ledgers, some of them propose decentralized ledgers. They vary in terms of the end users' experience and how much leeway is given to the individuals operating the system as to how to design the front end that the public gets to see. Whether those things look like individual coins or they look like accounts is one way that the field can differ. It can differ in terms of the amount of identification needed to make a transaction work through.

It can differ in their interfaces. Some of them propose to do it through smartphones, others propose to do it through plastic cards, other proposals allow for both. They can differ in terms of the extra features they offer. Some are proposing smart contracts as part of the arrangement, others are proposing that these things have offline functionality. That is, you don't have to be connected to the internet, or your provider, the company that you're buying something from, doesn't have to connect to the internet at the instance that they're being used.

With all of those different alternatives available, you also have a variety of alternatives as to why we want to be having this in the first place. Some of the motivations out there for proposing a CBDC are to, first of all, possibly improve on cash. There are various downsides to cash. Maybe we could do it better electronically. Or possibly, there are some really good things about cash, but cash is disappearing because of all these other alternatives out there and we want to make sure that one alternative, at least, continues to have some of the special functionalities of cash still in it. There's a big concern with the possibility that what we want to be doing is encouraging financial inclusion, and that for individuals who are unbanked, and in particular unable to take on the benefits of electronic technologies, we want to be able to have a means of payment that they can enjoy and join in as well.

You can have stories about why we want CBDC because we want to encourage the existing financial institutions to do a better job of innovating and providing new forms of payments. Or, you can say, "Well, I like what they're actually doing, but I just want to make sure it's coordinated and safe." Some arguments are in favor of central bank digital currencies for providing a way of interconnecting different proposed or existing alternative means of payments or making them safer by having a safe and intermediate asset that people could move to if necessary.

All of those arguments are for retail currencies. There are also some important arguments out there for wholesale currencies. We think that there are some problems with the way that wholesale settlements are made right now. Perhaps they could be improved by using a digital currency that links up with the digital technologies that are available right now for settlement in financial institutions and in financial markets.

There are other macroeconomic arguments. We want to be able to have better control over monetary policy. Maybe it would be better to do that by having an electronic currency through which the Fed could make its own monetary policies felt. We're worried that if people stopped using central bank money, policies would lose their force, and so we want to make sure that it's still out there.

In some countries, seigniorage is still a consideration. Not for the United States, maybe so much, but there are plenty of countries out there that worry about the continued seigniorage revenues that they're receiving in their central banks. Maybe they want to have a digital currency to make sure they don't lose out on those.

That's a large number of arguments out there. Because there's such a large number of arguments it's actually become very difficult to figure out why you're talking about it. What are the key reasons for having it? In research on this question in various places, I've seen that what's happened is something of the following form: central bankers get out there and start looking at the advantages that a digital currency could provide, and as a result of that they come up with, "Yes, it can provide a large variety of benefits, but when you talk about any particular problem and examine that carefully, it's not the simplest or most effective solution for that particular problem." You get a lot of conversations along the lines of, "Well, if you're going to provide a CBDC anyway, it could probably help on this dimension as well."

It becomes important instead to distinguish between the real motivations and the side benefits. You need to think, "Is the CBDC not just useful along these lines, but the right solution for any particular problem?" before you start tacking on the additional side benefits on top of that, and thinking seriously about what alternatives are available out there for providing the solutions to the problems that you really care about.

Secondly, the second result that comes out of this survey of the reasons is a suspicion that the key political motivation that central banks have for looking at CBDCs is a fear of irrelevancy. The worry is that the central bank is going to be losing connection to the payments system and losing, possibly even more dire, connections to the unit of account. The energy that has recently come into looking at central bank digital currencies partly comes from that fear of irrelevancy.

As justification for that argument, let me point to the difference between the attitude that's been taken in response to Bitcoin several years ago and the attitude that was taken in response to Libra, the proposal by Facebook much more recently, who are putting this together. When Bitcoin came into play, the attitude of central bankers was, "Oh, isn't that interesting. Aren't those cute little techies doing something neat there? This is something we should study, but there's no rush." The attitude towards Libra: "Oh, those guys are serious business guys. We need to think about it, because this really might make a difference. This really might take over our turf, take over our territory." I think that that's an important thing to understand when looking at these motivations.

With that introduction, let me talk quickly about the different examples I have. The first example I've got is Sweden. Sweden was very early to the game. If there are any Fins in the room, I'll have to admit that Finland now brags about the fact that they actually had the first CBDC. They had a working example of a prepaid card that was actually a CBDC working very well back in the 1990s, but let's leave that one aside. For the current round of discussion, Sweden was very early to the game, and they were very early to the game because of the extreme position they were in, where cash was disappearing rapidly in Sweden. They were one of the first for this to happen, and they are still outliers in the nonuse of cash.

At the time that that happened, everyone was very enthusiastic about the disappearance of cash. They thought this was a wonderful possibility because cash induced all sorts of antisocial activities, and so a cashless society would be a wonderful thing to have. That was the attitude at the very beginning. As time went on people began to understand that there were serious disadvantages from the loss of cash, and in fact the most vulnerable members of society were most dependent upon cash. If cash was going to disappear you better find some viable alternatives for them. That was the attitude that was available when the first reports came out of Sweden.

Currently in Sweden right now, there's still discussion going on between the government and the central banks about whether they should be having a central bank digital currency, but they are also doing proof of concepts right now on the same line as the Project Hamilton that is going on. Their arrangement has three interesting features that we want to think about. First of all, it's a pretty complicated arrangement. It allows for both the central bank and wallet providers out there, each dividing responsibilities among themselves. Secondly, the proposal that they are arranging really has very little privacy associated with it. Anonymity is sort of in the background of the story that they're telling, sort of treated as an afterthought. The third thing which needs to be compared with Project Hamilton is that there seems to be no serious study yet of how you scale this thing. How do you make throughput? One of the interesting things about the Project Hamilton study is that it focuses very much on is throughput possible. Can we make this thing big enough to be able to be used?

The second example is Canada. Canada was also early to the game, but their focus at the beginning was on wholesale CBDC. They put together a project called Project Jasper in which they looked to see whether a decentralized ledger, a token, could be used for doing wholesale payments. They looked at two particular cases. They looked first of all at the possibilities for using it for making interbank payments to substitute for their own ledger. They also looked at the possibility of using it to do settlement on stock markets or other financial markets. Their conclusions were, "Yeah, you can do this. Not only can you do it, it's really easy." They set it up really fast, their pilots, but they concluded there wasn't really any advantage from doing this, that these things work pretty well already, and there wasn't much that this was buying them that they didn't already have. They admitted that if you got to the point where there were lots and lots of different assets being traded on blockchains or other fancy arrangements, it might be helpful to have a digital currency on these arrangements as well to facilitate those. That was left as an open question, but the answer was, "If you take any individual market, it was not going to get you anywhere."

They also looked at retail CBDC and put together a position paper on that. They concluded that there was no immediate need for one, but that it was important to keep contingency planning going on. That sounds familiar to me. But they gave two interesting cases where it would become important to come in and jump into CBDC. The first was if cash really did disappear, and the second one was sort of a monetary sovereignty concern: suppose it's the case that some noncentral bank form of payment begins to dominate the market. We're going to need to respond to it, and that would be another argument for it. Until one of those circumstances arose, no, there was no reason to be out there actively putting the program together, just preparation for readiness. By the way, they're also putting together an arrangement with MIT to also look at feasibility along those same lines.

The third example: the Bahamas. The Bahamas are unusual because they are the one place that I've got in my examples where there actually is a CBDC working right now. It's been there since 2020. It has 20,000 active wallets. That may sound not worth worrying about, except that the Bahamas only has 400,000 people, so 20,000 wallets is a pretty good number out there. They have a very specific need that motivates their CBDC. Their need is motivated by geography. They've got a lot of remote islands which are outside of the reach of physical banking, where it is really expensive to get cash out to the people in those islands. It becomes very sensible to try to find an electronic alternative for it. Furthermore, they've got hurricanes, and if you've got hurricanes you're worried about things going down, so you want to have some electronic thing that's resilient enough to be able to handle emergencies.

With those two motivations, they put their Sand Dollar together. The way the Sand Dollar works is that individuals get wallets. Wallets come in various tiers. Your simplest tier of wallet, Tier One, has a maximum of $500 in it. Doesn't require a bank account; doesn't require any serious identification. The larger tier has a $5,000 limit, has to be linked to a bank account. Any extra monies that you get above the limit gets flowed over into your bank account. And it requires some identification. If you're a business, your wallet can be much larger than that but you have to have links that demonstrate your tax identification so records can be kept for it as well. It's implemented with either a phone or a card. Unfortunately, they discovered that this offline functionality, which was one of their main issues, was actually still a problem for them. They can't get it to work as they wanted, so there's still things that need to be fixed with their system, but it is up and running and a success for at least some of the things it was intended for.

China still is doing its projects as a proof of concept or a pilot project. However, a pilot project in China is much bigger than a running project in the Bahamas. The pilot project in China has approximately 1.5 million wallets going in it. They're being used in 10 different locales, including being used in the Winter Olympics. China has this very specific situation, which is very interesting because in China right now, nonbank electronic payments already dominate the payments system. Alipay, which is a payments system run by Ant Group, and WeChat Pay, run by Tencent, each have approximately a billion wallets out of a population of 1.5 billion. Although 1.5 million wallets sounds like a very impressive arrangement, you have to take it in context of the size of the country that we're talking about here.

In the pilot program, they're testing the payments processes and technologies, and testing add-on features. Add-on features include facial recognition, tap-and-go, and offline capabilities as well. It's currently got very small throughput, but their ambition is to have this throughput that are comparable to those of the big guys out there already.

How would you get money into the system: electronic cards, or mobile wallets? Either standalone, or that could be paid into the system through Alipay and WeChat, where you put your money in that you already had in one of those systems and move it over to the other. The interesting motivations they have besides the standard motivations that everyone else has are they're worried about the control that Alipay and WeChat have over the payments system. What can they do, in the PBOC [People's Bank of China], to catch up with that? There's also a suspicion that part of the motivation is to maintain social control; that this can be linked up easily to their social credit system.

Finally, another interesting part of the story in China is the international aspect. A variety of initiatives have been done in China to increase the importance of the RMB [renminbi] to a status that is comparable with the country's importance in international trade. We can think of the e-yuan [e-CNY] as an attempt to do precisely that. Along those lines, one of the most interesting possibilities for the e-yuan is to be connected with other central bank digital currencies and for it to be a way of facilitating trade internationally. Along those lines, there is a project going on by the Bank for International Settlements, experiments in linking of CBDCs for precisely such an arrangement.

The potentials of that are very great, so it should be noted that the PBOC has made an explicit policy of noninterference. That is, to say, "It is the case that we're not interested in using our CBDC, should it arise, to be running domestic transactions in other countries. Instead, the goal is to have an interface between our CBDC and other CBDCs to facilitate international payments." That's consistent with the design of the experiments that are being run within the BIS right now.

Okay, lots of examples there. The question is, do we have messages out of those for the United States? I think we do, and I think we have to separate how those messages work for, first of all, wholesale versus retail; and secondly, for domestic versus international uses. On the domestic retail front, let me point out that many of the rationales used by other countries simply don't apply in the United States. There's no groundswell against the inefficiency of domestic payments in this country. There's no danger of invasion from foreign systems in this country. Cash is being reduced, but there's actually no rapid disappearance. Even during COVID, cash remained remarkably resilient.

Now, it could be the case that a basic CBDC could be useful for providing access and act as a competitive spur, so there are legitimate arguments there for pushing through the CBDC. In that case, I have to go back to my original warning: you want to think about, having focused on those, what alternatives are out there that might do this more effectively. The alternatives that have to be considered seriously include opening up Fedwire to a broader class of financial institutions in order to allow for interoperability and increased incentives for innovation. Also, alleviating the barrier to entry problems that might be slowing things down. The bigger question, it seems to me, is: How will faster payments play out, and who will gain access to Fedwire?

On the wholesale front, lessons are different. On the wholesale front, it has been said, and I think it's correct, a CBDC is a solution looking for a problem. It's true that you might be able to improve the running of settlement systems by using smart contracts and other technological improvements, but if those systems remain in place those same technical improvements, it seems, could be achieved with payments provisions within the system rather than needing to impose one from outside.

Perhaps you have in mind something broader, where what you want to do is free all settlement systems in the country from the existing arrangements so that people can float their liquidity back and forth between one settlement arrangement and another settlement arrangement. That's very exciting, but it's very speculative. It's going to be a long time before things like that can sort themselves out, including the problems, the legal dangers, from trying to do automatic execution of smart contracts.

Those large, more radical suggestions, I think, are at this point, pie in the sky. The issues instead for settlement systems are the question of how the pipes are working, how efficiently they're run, more so than the question of what's the nature of the payments asset that's going to be flowing through those pipes.

Finally, for international payments, there are more legitimate arguments, it seems, for CBDC. CBDCs could certainly eliminate one level of the complexity at the interface. Although Canada's experience is cautionary that these systems work pretty well, it's still worth continuing those kinds of experiments, so the BIS work on this is extremely important.

At the retail level, you probably wouldn't need a CBDC to do this. You could use a regulated stablecoin or bank money itself to improve arrangements. CBDC doesn't become so necessary. At the wholesale level, it probably would be useful because it would give that little bit of extra security and safety that the thing that's going through the pipes is something that's guaranteed directly by a central bank.

I'm running over. I'll talk about privacy briefly. The attitude towards privacy has changed in the time that this research has been going on. It used to be that the only reason for having money, it was believed before, was doing all sorts of bad things like tax evasion, money laundering, and illegal purchases. Now, it's better understood that we want privacy for some very important, very good reasons: protections from harassment, and protections from failures of the payment service provider to provide the privacy, to provide the protection from harassment. That is, protection both at the direct level and at the indirect level.

Part of our desires for understanding CBDCs comes from trying to understand how privacy will change if CBDCs come into play, and that's a really complicated question, depending very much on the details of the CBDC in question. Notice that in the Bahamas example they were going to give privacy at different levels for different sizes or classes of wallets.

Part of the puzzle right now is the question of who is best able to provide privacy. Should privacy be provided by private or public institutions, and who does the consumer trust most for those kinds of privacy? As we've seen, attitudes towards privacy differ greatly among the countries that I have been describing to you. One important reason that might be out there for individual countries to be interested in researching CBDCs is to continue to have a seat at the table as the ground rules are put together for how privacy will be implemented in international retail arrangements.

Let me summarize: other countries' experiences provide a touch of realism, I think, to the US discussion of CBDC. There's a great tendency to talk in glittering generalities, so looking at what's actually done and what actually can be done helps to bring the discussion back down to what are we really interested in. Secondly, in the US, I would think that the best use cases are for a basic service that can be used for inclusion at a very basic level and possibly provide a convenient linkage between more sophisticated offerings by various private institutions. Finally, the rules about the pipes through which payments flow are probably going to be more important for most cases than the exact nature of the payment asset that's flowing through them. Thank you very much.

David Mills: Thanks Charlie, thanks Nellie, and thanks to the Atlanta Fed for including me in this conference. It's always a fun topic to be talking about, central bank digital currencies. What I'm going to talk about today in my prepared remarks is really where the Fed is at in terms of its thinking around central bank digital currencies, and also cover a lot of what is in the discussion paper that the Board of Governors released in January and is currently in a comment period process, at least until May 20. If you haven't started giving us feedback, now is the time to get going so we can maybe start working on it at the break.

Let me provide a bit of an overview of where the Fed is at in terms of our recent paper, "Money and Payments: The US Dollar in the Age of Digital Transformation." It's about CBDCs, but it's even a little bit more about that. It's about the future of money, and really thinking about the future of money and payments and whether or not the dollar needs to evolve. This is an existential question. It's a big picture question, and as we discussed, it's also quite an abstract question because we have a lot of different issues and topics to be thinking about.

What we're trying to think about as an institution is to begin a dialogue or conversation with a variety of stakeholders, both domestically and internationally, both in the public and the private sector, general citizens as well as institutions, to dissect and understand what is a range of potentially complex, interesting, but also very important issues to think about when it comes to central bank digital currencies.

Chair [Jerome] Powell has often been on the record as saying the US central bank, or the Fed, is very deliberative; this is not something that we would rush to do. It's something that we need to be deliberative about and really think about. If this were a path, and we have no decision to move ahead at this point in time, to make about central bank digital currencies, we would have to ensure that we do it right. That really means recognizing that there are a range of views and perspectives that need to be thought about when thinking about the topic of adding a new form factor for the US dollar.

We've seen some definitions of central bank digital currencies. Why not a third? We define in the discussion paper a central bank digital currency as simply a digital liability of the central bank that is widely available to the general public. So this is this general purpose idea. In this respect, it's analogous to a digital form of paper money, but it's a little bit more complicated than that because what we're not thinking about is replicating physical currency as it is, with all of its benefits and problems and challenges, in a world that is completely anonymous. That's not exactly what we're thinking about doing. At the other extreme, we're not thinking about replacing the banking sector by offering accounts to all citizens in the US. Central bank digital currency is something in between, and that's why it's also a bit hard to unpack here. That's kind of where we think about it, in terms of that it's a new type of form factor for the dollar. It is a liability of the central bank, but beyond that there's a bit of ambiguity in terms of what this could be and there's a range of possibilities on what this could particularly look like.

As I mentioned, we don't have any policy outcome. We're being very deliberative and very agnostic at this point in time and beginning to have this conversation through this discussion paper and subsequent comment period.

What does the paper try to cover? It covers some things that both Nellie and Charlie already kind of touched upon, so I'm just going to give a bit of a high-level profile on some of the potential benefits and uses. One of the things that we want to try to consider or emphasize is thinking about if is there some basic foundational layer that would be sort of a public utility. Nellie had on her slide what's the market failure here, or the public good nature of this. Is there this foundational layer that, building on some new technology and allowing for a new form factor for central bank digital liabilities, would be able to serve the foundation for the monetary and financial sector, moving ahead over the next several years and decades?

That's one big picture question: are there some merits in some new technology over which there is a central bank role to provide a new liability in digital form with potentially new types of infrastructure than what we already have today? That's a big picture question, and what motivates that a bit is the fact that change is happening around us. Payments and money is actually not a very static ecosystem. Nellie's slide, I think, points to not just the various form factors we have today, but it's also so much of an evolution of how money and payments infrastructure has evolved over time, and it doesn't sit still. Indeed, the Fed is currently building out a real-time, instant payments, retail payment service for banks, called the FedNow Service, that will also add a new layer of functionality to an already digital and vibrant payments system. Can we do more? That's a question. Charlie had mentioned also cross-border payments. Is there some aspect of introducing digital forms of the dollar that would actually improve on cross-border payments? I think we'll talk a little bit more substantively about that in the Q&A, so I'll just leave it at that.

We've mentioned the importance of the dollar internationally; it is obviously a reserve currency. It is a big percentage of cross-border and international payments, so we have a responsibility to think a bit about how anything that the Fed would do in terms of a central bank digital currency factors in the importance of that feature as well.

It was also mentioned earlier today, financial inclusion. Is there a way to provide additional access to consumers who may be underserved, at least in the digital payments ecosystem? Is offering a digital liability of the central bank a gateway to broader access to financial products and services?

Finally, and Charlie mentioned Sweden, is there a need for some public access to a safe, central bank money should physical currency decline? Now, the Fed is in no business of eliminating physical cash or physical currency. They're thinking of central bank digital currency as additive to existing ways that people pay and existing form factors of money, not as a replacement, which is an important tenet to keep in mind but also a challenge when you think about what's new and what's different to offer.

There are scenarios where cash could rapidly disappear because the momentum for digital payments and digital transactions actually really de-emphasizes the use of physical currency, at least on a day-to-day basis. Does that mean that consumers and households don't have alternatives, like they do with physical cash today, that they could practically use in a range of transactions?

The paper then also goes a bit more into some of the key risks and policy considerations. There's quite a bit, and these are all interrelated, which adds a little bit to some of the complexity around central bank digital currencies. First and foremost, if you're issuing a new type of digital central bank liability that is widely accessible, how does that impact the financial market infrastructure? Does the market structure for banking change? Do we have to worry about disintermediation of the financial sector, banking sector?

What does that have for implications for the broader credit market in the steady state, where people are holding more central bank liabilities, and maybe holding less deposits? How do we think of it in regard to that? Where would bank credit evolve, or how would it be replaced by other sources of credit for the macro economy? These microeconomic impact questions around market structure have macroeconomic implications. Trying to sort out and think through some of those issues are important.

Obviously, another key issue not just tied to disintermediation in the banking sector, but also potential flights to quality for a central bank digital currency: does that have impact on the safety and stability of the financial system? Are there negative effects? There could also be positive effects as people have already referenced in the panel today around stablecoins. If stablecoins are still outside the regulatory perimeter and people are relying on them for a significant part of financial transactions, does the CBDC help provide some more stability and foundation for those types of transactions, if the regulatory perimeter issues can't be solved? So, another important question.

Charlie just mentioned consumer privacy. How do we balance that with transparency necessary to deter criminal activity? So again, can we use technology to provide a layer of privacy through transactions, but maintain the identity verification in ways that also will help detect and prevent illicit financial transactions?

Finally, I think Raphael [Bostic] just mentioned earlier, what keeps him up at night is around cyber security and operational resiliency. These are all important issues that one needs to think about if you are to build an alternative infrastructure for a central bank digital currency and really be able to design something that would be robust to what is a continuing threat space in the cyber world.

Out of that, as I mentioned, the Fed is fairly agnostic about any sort of specific principles. One, we're not in a position to make any decision about a central bank digital currency. I mentioned that; I'll probably mention it three or four more times, if you'd like. Certainly, we do have a lot of questions that we're interested in about its viability and its use cases, and how to think about some of the risks and challenges that are identified. If we were to design it, we do have a few simple principles or guardrails around what a central bank digital currency could look like. That's what these on this slide are, here. I mentioned privacy protection. I also mentioned identity verification. They're a bit at tension, potentially, so we have to think about how you might thread that needle. Those are two very foundational, important issues. Charlie mentioned privacy from whom. Again, I think a lot of our starting point is, privacy from the government, privacy from the Fed. This is not something that the Fed envisions being able to identify all of a consumers' transactions.

How can we leverage new technology to introduce privacy for consumers in transactions? I would say another example that's a bit more tangible for the payments space is today's digital payments ecosystem requires a lot of information in the transaction that often leads to data breaches and other types of stolen identities and synthetic identities. Can an infrastructure replace some of the requirements there and put them in safer vaults where less information, at least to execute the transaction, is necessary? It's something for us to be thinking about in thinking about what's possible and feasible here.

The other two design principles are thinking about the intersection of central bank digital currency in today's financial ecosystem. We would argue that another design principle is that a central bank digital currency be intermediated; that means that there would be a distribution model that would include eligible financial institutions. Today that, of course, is banks, but there is this question about whether other eligible financial institutions, if adequately regulated, could also provide a limited set of circumstances. Charlie mentioned the Bahamas. The Bahamas have a narrow bank wallet-type of arrangement that is a gateway to thinking about inclusion aspects, so that might be a way to think about that as well in the US case, at least for some purposes.

Finally, transferable. Again thinking about how does a central bank digital currency actually leverage not just new infrastructure, perhaps, but also could it be transferred on legacy infrastructure. This may be pie in the sky, because interoperability, interconnection, the workings of different infrastructure technologies, is a difficult problem. It's a difficult problem domestically, and it's difficult problem when we think about CBDCs in the cross-border space. This is not a trivial principle to solve, but one where we're hoping to get a lot of feedback on ways in which that needle could be threaded.

Very quickly, the work ahead. We have our comment period; it ends on May 20. We're unpacking and doing a lot of work within the Fed on the issues around the policy questions that we've talked about, thinking a bit more on that. We're doing a lot of stakeholder outreach and engagement around central bank digital currency, so we're holding sessions where we're talking about it, answering questions, socializing the process for comments, et cetera. I hear there's one at lunch where we could have more of a dialogue on this paper. Also, we're engaging a lot on technology, research, and experimentation, so there's a number of projects. Charlie mentioned Project Hamilton. We have work going on collaboratively with the Bank for International Settlements through our New York Federal Reserve Bank, and we have other pockets around the system that we're contemplating and thinking through to leverage and unpack the possibilities of technology to help frame and think about what is possible here and what is not possible.

Finally, Nellie mentioned the executive order. I would add that the Fed's role in the executive order when it comes to central bank digital currencies is we're being consulted. Obviously, there's a placeholder that encourages the Fed to continue doing its research and work, which hopefully you're convinced from the previous slide that we're doing that, but we don't have any direct or specific role related to central bank digital currencies through the executive order, because of course we're an independent central bank, so there's a bit of nuance there that's important to know. We are certainly engaged and working collaboratively with our colleagues at various agencies on a range of the topics within the executive order.

With that, that's my remarks, Thank you.

Paul Kupiec: Good morning. I'd like to thank the Federal Reserve Bank of Atlanta for inviting me, and for allowing me to participate in this panel this morning. Should the US issue a central bank digital currency? The question brings to mind a famous exchange that took place at the Federal Reserve Board Monday morning briefing sometime in the 1970s. At that time, Ed Ettin, a legendary Federal Reserve Board officer, was deputy director of the Division of Monetary Affairs. He was briefing Chairman Arthur Burns during a period known in Fed history as the Great Inflation. According to Ed himself, he and Chairman Burns did not always see eye-to-eye on issues of monetary policy. At some point in the briefing, Chairman Burns asked Mr. Ettin a particularly difficult question. Ed's response to Chairman Burns was, "Mr. Chairman, there's a short answer and a long answer to your question." When Chairman Burns asked for the short answer, Ed responded, "I don't know." Don't you wish more of us economists could be so to the point?

With that as background, let me give you both a short answer and a long answer to today's question. Should the US issue a central bank digital currency? My short answer is no, and now I'm going to go into my long answer.

To begin with, why are we even talking about this question today? We've already reviewed there are many reasons: the popularity of Bitcoin, the launch, or potential launch, of Facebook's Libra, subsequently known as DM, which ultimately failed, the whole growth of private stablecoins other than DM, and then the fact that it could be problematic if another foreign central bank issues a digital currency first and it becomes very popular. It might eat into the special privilege the US dollar enjoys.

A Federal Reserve digital currency, should it be issued, would be a direct liability of the Federal Reserve. As such, it would be free of default risk, meaning that you could take your one Federal reserve digital currency dollar and always redeem it for a $1 Federal Reserve note. A Federal Reserve digital currency would be the ultimate safe asset and a magnet for investors seeking safety.

While the idea of a Fed digital dollar has only recently been elevated to the front pages of the Wall Street Journal, the idea has actually been around awhile. In August 2017, Jamie McAndrews and his colleagues formed a limited-purpose Connecticut bank called TNB, or The Narrow Bank. TNB's business model was simple. It would only take large deposits, wholesale deposits, from corporate treasurers, money market mutual funds, institutional investors, and it would invest them in a master account at the Fed. TNB deposits at the Fed would basically get paid the interest rate on reserves. TNB's deposits would not have deposit insurance. TNB would keep a portion of the interest and pass the rest onto its depositors. Deposits at the TNB Bank would be, in all but name, Federal Reserve digital currency.

TNB's application for a Fed Master Account was slow-walked by the Federal Reserve Bank of New York. FRBNY launched an especially vigorous and time-intensive analysis of the risk that TNB might pose to the system. When ultimately the Federal Reserve Bank of New York could not come up with a reason to deny TNB a master account, the Federal Reserve Board reportedly intervened to prohibit the FRBNY from opening a master account. The Federal Reserve Board argued that TNB deposits, which are Federal Reserve digital currency in all but name, could pose a systemic risk to the banking system.

There is at least one other bank, and maybe two, that are chartered under Wyoming law right now who currently have applications for master accounts that have been in play for almost two years with the Federal Reserve Bank of Kansas City. There are at least two I know of.

President Biden's March executive order unleashes a whole of government approach for assessing the risks associated with crypto assets, including private stablecoins. It mandates that the Federal Reserve, US Treasury, and US Department of Justice report on the legality of, and the risks and benefits associated with, the Federal Reserve digital currency. Will this whole of government effort convince the Federal Reserve Board to reverse its TNB decisions?

If it does, politics likely could play a role in shaping the design of the Federal Reserve digital currency that's issued. In 2020, Senator Sherrod Brown pushed hard to get his Banking Act for All included in the CARES Act COVID relief legislation. The bill would have created a Federal Reserve digital currency that provided free banking services for everyone. At the time, the Senate was controlled by the Republicans, and his efforts failed. Senator Brown's bill would have required Federal Reserve district banks and member banks to offer a new type of public digital currency, free of charge to the public. These digital wallets, called Fed Accounts, would hold Fed digital dollars, pay interest, and provide all the services typically associated with a full-service commercial bank checking account: a debit card, ATM access, and electronic bill paying services with no minimum or maximum balances. The bill would have required large banks to absorb the cost of offering Fed Accounts, while smaller banks, banks smaller than $10 billion in assets, would have had their operating costs reimbursed by the Fed.

More recently, congressman Tom Emmer introduced a bill that would prohibit Federal Reserve banks from offering products or services to individuals, thereby precluding Fed Accounts. In his view, this restriction would ensure that public digital currency, should it be issued, would provide the privacy services that mimic private stablecoins. Despite Congressman Emmer's revealed preferences regarding Federal Reserve digital currency design, I don't think his bill would preclude digital currency being issued in an intermediated way, as has been discussed earlier, by the Federal Reserve. This form would not be traded on the internet but could instead probably use the existing banking payments system.

The design of Federal Reserve digital currency is perhaps the most important consideration in weighing the cost-benefit considerations surrounding its issuance. Will it pay interest? Will transactions be processed on a distributed public ledger, or processed centrally on a system controlled by the Fed? Will politicians succeed in attaching their preferred equity and inclusion features in the Federal Reserve digital currency design? My best guess is, should the Fed issue Federal Reserve digital currency, it would likely use insured depository institutions and other licensed financial firms as intermediaries to hold and manage Federal Reserve digital currency accounts. Federal Reserve digital currency payments would likely clear and settle using a new system built and centrally managed by the Federal Reserve system in a manner similar to the way checks and ACH transactions clear and settle today.

In my opinion, it is highly unlikely that Federal Reserve digital currency transactions would be processed on a public distributed ledger. Whether or not the design would include equity and inclusion features is anybody's guess. If Federal Reserve digital currency takes this form, it will not be a substitute for private stablecoins. Private stablecoins are a competing form of digital money that is purchased and traded using the internet. To date, private stablecoins have not achieved universal acceptance as a means of payment, and their growth mostly reflects their use in facilitating the trading of other digital assets like Bitcoin.

Private stablecoins are a new solution to a very old problem: how to find a mutually acceptable way to pay someone who is physically distant, who you do not know, and who does not know or trust you. In ancient times, you could travel and carry metal specie, or the king's minted coins, but you might face the risk of being robbed along the way. In 1099, Jerusalem fell to the Christian crusaders. Thereafter, Christian pilgrims across Europe visited the Holy Land, but many of them were robbed and killed as they journeyed through Muslim-controlled territories. Around 1120, a French knight named Hugues de Payens founded a military order called the Poor Knights of the Temple of King Solomon, or the Knights Templar, they were later known as. The Knights Templar were warrior monks, bankers, pirates, and a religious order that reported directly to the pope. They had headquarters on Jerusalem's sacred Temple Mount and pledged to protect Christian visitors to the city.

To facilitate pilgrimage to the Holy Land, the Knights Templar accepted specie and coin and issued pilgrims paper bills that could be exchanged for money at preceptories from England throughout Christian Europe to the Holy Land. To prevent fraud, the Knights reportedly developed a system of codes for the safe passing of information, and pilgrims' deposits and withdrawals from preceptories were secured on paper bills using coded ciphers. In essence, the Knights invented the travelers cheque. Other historical accounts argue that these pious monks actually stole the idea from their Muslim adversaries. Proponents of this theory claim that a similar system existed in the Fifth Abbasid caliph, which would have put travelers' cheques in use from about the year 790. My guess is that those folks learned it from the Greeks or Romans earlier, but I haven't been able to confirm that.

A little short aside here: travelers cheques are a kind of interesting feature. I've been a bank regulator for a long time, and I never really looked at issues with travelers cheques. Travelers cheques are really treated like letters of credit under the international code. They're not deposit instruments. They're not protected by deposit insurance. They don't have to be issued by banks. In fact, money orders are a close cousin of travelers cheques. They're just a different form of a letter of credit. You can get a money order at the post office. You've been able to do that for a long time. Travelers cheques and money orders actually clear through the Fed system that clears checks. They clear like checks, but they're not checks. They're not deposit accounts. They're prepaid products, is what they're called. If you look on the American Express annual report, you don't find a column entry for travelers cheques. It's listed in prepaid products; it's a prepaid product. Very interesting: travelers cheques are included in M1. They're part of the M1 money supply. I don't know that money orders are. I don't think bank letters of credit are. But then, if stablecoins are really just the modern incarnation of travelers cheques, should they be included in M1? I'm going to leave that. There's a bunch of real monetary guys out here that know the aggregates. I'm going to leave that question out there, but it's an interesting one.

Modern stablecoins are digital assets designed to maintain a stable value relative to a reference currency, like the US dollar, or commodity, like gold. Many stablecoins attempt to maintain their value by investing their dollar proceeds from newly issued stablecoins and high quality, short-term, liquid dollar-denominated assets of equivalent value, held by the stablecoins' sponsor as a reserve that can be used to stabilize the coins' market value. There are other versions of stablecoins that are really an entirely separate breed. They hold crypto assets as reserves, or they use algorithmic arbitrage trading to maintain parity with the dollar. It's unclear to me why those other types of coins should be recognized as a legitimate means of payment rather than a risky security.

Stablecoin transactions are processed using a public distributed ledger system. You've heard about that. Agents compete to earn rewards for processing stablecoin transactions, but there are a lot of different stablecoins out there now, they're processed on different ledgers, and they're not interoperable. Thus far, private stablecoins have not been issued by any insured depository institution that I'm aware of. Rather, they have been issued by entities that are either unlicensed, licensed as state-regulated money transfer agents, or licensed as limited purpose trust companies.

What can private stablecoins do that a Federal Reserve digital currency couldn't do better? To answer, let me borrow a famous line from the movie The Graduate. You may remember when Dustin Hoffman is at the party and Mr. McGuire takes him out of the party to the sun, and he says to him, "I want to say two words to you. Two words; are you listening? Smart contracts. There's a great future in smart contracts."

The next iteration of the internet, so called Web 3.0, will have the ability to automatically execute transactions using digital currency. Imagine a future where your refrigerator monitors its contents, compares them with a list of items you specify and automatically orders from your favorite grocer, who delivers the goods to your door. Your refrigerator will pay for the order automatically with digital currency. This is the Buck Rogers world of Web 3.0. It is unclear to me why your refrigerator could not pay for it using a credit or debit card, but cryptocurrency developers think in terms of a public ledger payment system when they design the smart contracts that will automatically restock your refrigerator. If Web 3.0 requires payments to be processed on a public distributed ledger, then if you're holding Federal Reserve digital currency, you're going to have to go to the grocery store. While the crypto industry proponents argue that a public distributed ledger payment system is required to facilitate smart contracts and Web 3.0 functions, I'm not sure why that must be true. However, what is clear is that the public distributed ledger systems have been the key factor driving smart contract innovation.

Without any viable alternative at present, it seems important that the public distributed ledger payment system continue to exist. The case against Federal Reserve digital currency goes beyond just the need to fuel smart contracts. Unless Federal Reserve digital currency holdings are limited by law or regulation in a crisis, their status as the ultimate safe asset will attract large balance transfers from banks and money funds. In other words, they'll create a new, formidable liquidity risk for the financial sector, similar to the risk that the Federal Reserve Board cited to deny TNB a Fed Master Account. Federal Reserve digital currency has a potential downside even in normal times. Purchases have to be funded either from bank deposits or money market fund withdrawals, and there could be a drain on intermediaries' funding and have an effect on credit costs and availability in the economy. They are also going to require the Fed to perpetually maintain a larger balance sheet than it otherwise might. Bank deposits are an alternative form of digital currency, but deposits over the $250,000 federal insurance limit are technically at risk should a bank fail. Moreover, deposit payments clear and settle over systems centrally controlled by banks and the Federal Reserve. At least today, these systems will not support the use of smart contracts being developed by the private stablecoins in the private stablecoin space.

Now we're getting to the shorter part of my long answer. So, what will the digital currency of the future look like? Will it be private stablecoins, Federal Reserve digital currency, or both? My guess is that the Web 3.0 will run on a hybrid digital currency that includes a new type of bank deposit, as well as private stablecoins. A simple solution that solves the tension between private stablecoins and Federal Reserve digital currency issuance without requiring innumerable government studies, new regulations, or Federal Reserve digital currency issuance is as follows: banks and firms licensed to issue private stablecoins form a consortium, a jointly owned payments system processor. That processor develops and runs a public ledger payments system or platform that can be used by private stablecoin issuers and banks alike.

Such an arrangement mirrors the way that credit, debit, ATM, and ACH processing systems were developed in the past. That's exactly the way card interchange systems and things like that, ATM and debit systems, were developed over the last 35 to 40 years. The energy-efficient public ledger they will develop will use some type of secure proof of stake system, where banks and qualified nonfinancial institutions will compete to process transactions. Like other payment systems, the Fed would have oversight powers and could require that proof of stake processors satisfy certain minimum requirements.

Banks would offer a new, tokenized deposit account. There's no regulation I'm aware of that prevents insured depository institutions from developing tokenized, insured deposit accounts that can be traded on this new payments platform. These fractional reserve deposits will be a new type of checking account. We already have bank capital liquidity and other regulations in place to manage the associated risks, so I don't think we'd need to do a whole lot of extra studies and regulations to make this happen.

Similarly, licensed private payment stablecoin issuers, like those envisioned in the stablecoin TRUST Act, introduced in draft form by Senator [Pat] Toomey recently, would create tokens that can use this common payments platform, which would ensure interoperability, and it would create the competition necessary to ensure that the public accrues the benefits of this new form of digital money. With this system, there would be no need for Federal Reserve digital currency, or the dozens of new government reports and resulting regulations that the President's March executive order will surely produce. I thank you very much for the time, and I'm sure we're going to have an interesting discussion. Thanks, Nellie.

Liang: Thank you for all those presentations. That was great. We've had a whole slew of questions come in, so I'm just going to start with them. A number of them are about what the new payments system in the future should look like; a number of them turn to some monetary policy considerations. Let me just start with what the new system looks like. I think this is for everyone here. "If the US had a more established framework for regulating and supervising digital currency, in this case, let's say stablecoins (this is exactly where Paul led us) does the central bank need to issue its own?" You can say "no," but if you could just elaborate a little more on that, and then we could do Dave and Charles, that would be great.

Kupiec: When I started doing stablecoins, and when all this innovation came about ... I'll admit I did bank regulation for many, many years, but I didn't really know anything about the payments system. Pat Parkinson was the payments system guru at the board, and that was his area. The rest of us dealt with credit risk, liquidity risk, stress testing.

I've read a bunch of stuff about payments systems, the history of them, where they came from, the development. It was really fascinating how it came about, and it started in the late '50s into the '60s, and how it developed. It struck me that this is the same revolution we went through when they developed card payment systems and debit card payment systems. It's just a different way to do payments. There's no reason why the banking system couldn't be involved. The way they developed over time, banks got together, they formed a group to create rules and oversights and protocols, and they developed computer systems and ran them.

Somebody always gets paid when you transfer money from A to B. There's always somebody that gets paid there, right? It's never free; payments cost money. They'll have to figure out how to run that. Bitcoin costs money. The Ethereum network costs money; you have to put in gas. It's not like the internet-traded digital coins are free. If you want to use them, somebody somewhere pays to process them and use them. It just struck me that, really, we're just at this road and this just makes a lot of sense. Why not have checking accounts that are digitally tokenized? I'm surprised we haven't seen that yet. But maybe you have, and I just haven't.

Mills: I think one of the biggest challenges that we face at the Fed is really trying to understand what the future state of money and payments is going to look like, and how it evolves. There are different ways in which the future can evolve and manifest itself, and I certainly think that one future, with the world of tokenized bank deposits, makes a lot of sense.

I think it also begs the question: do we need a wholesale infrastructure on top of that to tokenize dollars? I don't know. I think these are all the kinds of questions that we're posing, because there's states where that could be done. If we can't get the regulatory clarity we need, what are the alternatives that might emerge? And, what does doing nothing do? Does this evolve in a way that stays outside the regulatory perimeter, and we find ourselves in a particularly challenging financial stability type of concern?

I think these are all plausible outcomes; and what's really hard is to actually identify what path would evolve, and how we can steer to the right path or the right set of paths. I certainly think that what was being proposed is a very plausible potential outcome.

Kahn: I agree. Remember that we had about 14 different arguments as to why we would want CBDC. If we're focusing on the ones about being able to innovate a particular new payment, that sounds very reasonable, very plausible that we could do that particular agreement without a CBDC. Now let's talk about the other 13 possibilities as well.

Liang: Yes. So, maybe just to clarify, and just building this out: "How do you view CBDC versus FedNow, with broader access for the payments system?"

Kahn: I gave mine on that one already.

Liang: Okay; got it. I just wanted to emphasize that point. I think it comes back, at times.

Kupiec: There's a lot of progress made on faster payments. FedNow is another system to make that happen. NACHA [National Automated Clearing House Association] has a faster system. You pay a little extra, get faster payments. You want it faster, somebody's going to charge you a little more for it. That's okay.

I think one of the issues with direct smart contracts and digital dollar money is, when I talk to some people ... and I've never managed a payments system in a large corporation or anything like that ... they argue that they would like to have direct ability to program payments instead of having to tell the bank to make this payment on this day, and have an interface with the bank and trust the bank to do it and follow up. Right now, our payments system is kind of clunky that way, and that if you want to use the payments system, you've got to give your bank a set of instructions, you have to make sure that they actually carry them out, and then you have to follow through whether everything clears the way you think it's going to clear. The people working on the new transaction, they like the idea that you can go right to the internet, you can program it up. You guys may be thinking a lot about the risks there. I mean, is there somebody in between to stop it? But that's kind of what I hear from people that are working on this stuff.

Mills: We get this question all the time. How does this add anything over FedNow? I think the key answer, and I don't know what the answer is, but the key answer would lie under what sort of new technology that's not being leveraged by FedNow would bring. I think it's whether there is some value to issuing tokens, and then those tokens would have some kind of programmability, smart contracting capability, that provides enhanced payments services that you couldn't necessarily envision with today's simple debit and credit model that we have for FedNow and other payments.

It seems like that's one of the key aspects to try to understand, whether that's technologically feasible and can that be technologically feasible within the right guard rails, from a safety and soundness perspective. I think that's part of the way to try to at least begin to understand whether there's an answer to that question in the positive direction.

Liang: Let me turn to the issue of privacy versus security. How do we manage that tension? One specific question is, "Would there need to be a government-issued digital ID if we were to go to something like a CBDC?" Or, "What are the possible ways to address privacy concerns?"

Kahn: The different examples that are out there run the gamut. It is indeed possible to put together a system in which all uses of these things are sufficiently limited so that you don't worry about the privacy. There can't be more than $500 in there, or you go down to where there can't be more than $100. You can imagine it just being a way to make a change, or something like that, where you just say, "Privacy is not a consideration." Make it clunky enough at large values that you don't have to worry about people aggregating the thing into a bunch of different ... it becomes as inconvenient to try to amass 75,000 different cards as it would have been to try mass 75,000 different $100 bills.

You can build complete privacy in that way. The question, of course, is whether you want to do that, whether you want to allow your banks to do that, whether you trust them or not with the handling of that. You wouldn't need an ID, but for plenty of the designs that are proposed it doesn't make sense without sufficient identification, which either means linking to existing bank accounts, or for unbanked individuals to have identification required. All these kinds of designs are possible.

Mills: I think one of the reasons that the Fed's discussion paper highlights the importance of intermediate models is because intermediaries do this reasonably well, at least to some extent, in terms of providing a layer of privacy, at least privacy from government, again, getting back to who are you not sharing your information with, yet also being able to provide compliance around illicit financial activity.

That's kind of a starting point. The bigger question is, can you do more? The question then is on the intersection of technology and the rule of law and being very clear about how you tie identity verification on the one hand through intermediaries and the ability to transact with very little information that is stealable, compromise-able, from a lot of parties. Maybe there's some technology through tokenization and cryptography that enables at least certain minimum amounts of information to be transferred, that could have reduction in data breaches, reduction in other vulnerabilities. Would a national identity system help? I mean, one can imagine. Again, the details would be, how do you create such a directory with the assurances of protecting the privacy of citizens? Obviously, that gets concentrated in a few places. How it's designed would also matter.

Kahn: But you don't need it for most of the things you want to do.

Mills: You may not need it.

Kupiec: Well, in an intermediated model, we have anti-money laundering rules in place now. Financial institutions have to know your customer. They have reporting requirements for suspicious transactions. Travelers cheques and money orders actually have anti-money laundering rules in place, too. I think you're not allowed to sell more than $3,000 in travelers cheques or something a month. It's a smaller amount, so those are already in place. If you digitized accounts, it might create a new problem in that now, you might have one bank account, you might have two bank accounts. You're not going to have 10 bank accounts because it's just too hard to keep track of it all. But if it were digitized and digital money, one, that would mean it would be easy to get insurance for more than $250,000. You just spread it out across all the banks, and that actually might be a good thing for smaller banks. They would maybe attract deposits that wouldn't be there, that end up in JPMorgan or Bank of America. I don't know, I'm just throwing that out. It might be.

Kahn: The new version of a SLIP account.

Kupiec: Yes, something like that, but it might be harder to track suspicious transactions. If I've got accounts everywhere, and can program and use them, then somehow you've got to have some way to know that I'm spending money out of six different accounts at once if I'm doing things I'm not supposed to be doing. There are people in the crypto world that seem to think, "Okay, if it's intermediated, know your customer, we can deal with that." But there's other people in the crypto world that are like, "No way; it's got to be just as private as me using $20 bills. They can't know who I am at all." Which is kind of strange, because the blockchain...

Liang: It's a public blockchain.

Kupiec: You can actually figure out who owns what. We've seen that in a lot of high-profile stories recently. I'm no expert on privacy, but there are definitely a lot of issues there.

Liang: I will just comment, with private, even stablecoins now, which are basically not regulated prudentially, they are subject to BSA/AML [Bank Secrecy Act/Anti-Money Laundering] requirements at the state level. All the money transmitter licenses that they operate under have BSA/AML.

Can we turn to cross-border for a minute, domestic versus cross-border? e-CNY is offered domestically, maybe to some travelers who came for the Olympics but as I understand it most CBDCs are being considered for domestic. That's actually fairly common for payment systems. They're not meant to be interoperable globally. This seems to raise a whole slew of questions. Anytime anyone talks about CBDC, you start talking about interoperability across countries. How important is that? What was the new system? Do you want this emBridge thing? Just some thoughts on interoperability across countries would be useful.

Kahn: It's one sticking point. If I'm going to say, where might there be complaints about a system that exists, I think that's probably the one where the most pain arises for many people in the United States, is trying to send remittances across the border, or tourism, or whatever.

Liang: The issue is efficiency of payments and not some other value about national security or global financial leadership.

Kahn: Other countries would be worried about that. To me, that seems like a fairly remote worry for the United States. Maybe I'm naive about that.

Mills: Yes, I think when it comes to cross-border the G20 as a whole has been very focused on trying to make improvements on cross-border payments. I think the Facebook DM, previously Libra, had been a bit of a call to arms across the G20 community to think a bit about some of the pain points in cross-border. They've been longstanding, and there's no immediate or one-size-fits-all solution that ultimately, you say, "Aha; everything is going to work more smoothly."

There are these issues that the G20 is grappling with around expanding hours to be more 24x7x365, considering access points, access to international systems, linking wholesale systems. These are longstanding questions that have solid policy reasons why they're not easily connected, easily interoperated.

When people say CBDC is an answer to all this because it's a greenfield, it's not immediately obvious to me that the problem is with technology. It's more that there are reasons, geopolitical reasons, counterparty-risk reasons, financial stability and prudential concerns, that might just say, "Hey, we're not going to say, just because this is a greenfield set of technology, that we're going to be able to solve cross-border payments problems."

There are probably things around the edges where a central bank digital currency and infrastructure ... when we're talking about infrastructure, the infrastructure could involve tokenized deposits, or whatever it might be. There may be infrastructure enhancements to think of that might lead to better connections without necessarily having to solve what would be a complex diplomatic problem, in my view, of making progress.

Liang: Let me just conclude with one last question, just to get to a little bit of the monetary policy issues. So, central bank digital currency could offer interest, or not; that's a choice. I'll be interested in that. And the ability to lower the rate and basically lower the effective bound... is that an advantage or a disadvantage of CBDC?

Charles, why don't we start with you?

Kahn: If the theory of this is fascinating, I remain absolutely unconvinced that in practice the Federal Reserve would ever put a negative interest rate on anything. I don't see why we're even talking about it, to be honest with you. [laughter] Sorry about that.

Liang: Well, that's okay; it made it easy for David. Now he doesn't have to say anything.

Mills: It is interesting because a lot of the academic literature focuses on this very set of questions. But again, I think in practice this is not just an efficiency story, it's a political story. I think that there are reasons to think very carefully about those types of interest rate environments here.

I will say, one of the tensions you get with central bank digital currency, whether you pay interest or not, is the relative spread between it and commercial bank deposits, right? You have to think about what that does from a disintermediation perspective and the ability to do broader monetary policy implementation. These are again points to difficult questions we need to think about when you think about CBDC.

Kahn: Do we make it more currency or more banking? How far do we slide one way or the other?

Liang: Right. I think with that, we want to thank the paper presenter and discussants. We really appreciate the audience questions. Thank you very much.