George Selgin on Bitcoin and the Future of CBDCs

The future of a US CBDC might involve fintechs having access to the Fed Master Account.

George Selgin is the director emeritus of the Cato Institute’s Center for Monetary and Financial Alternatives and is a returning guest to Macro Musings. George rejoins David on the podcast to discuss cryptocurrency, stable coins, CBDCs, and a push for a higher inflation target. Specifically, George and David discuss the category of ‘synthetic commodity money’ and how bitcoin is a potential example, the current state of Bitcoin amidst El Salvador’s transition to Bitcoin as its legal tender, the role of fintechs in the potential future of a Fed central bank digital currency, and much more.

Read the full episode transcript:

Note: While transcripts are lightly edited, they are not rigorously proofed for accuracy. If you notice an error, please reach out to [email protected].

Beckworth: George, welcome back to the show.

Selgin: It's always a pleasure, David. Thank you.

Beckworth: Well, you are the reigning champ in terms of the number of appearances on this show. So you are now, I believe, doing show number nine. Setting some kind of record, George, on Macro Musings. Just for what it's worth, you're the second person I have now recorded within studio. So I'm actually looking at George face-to-face. I don't see him on a Zoom screen, which I normally have been doing. Hopefully we can do more of these, though some of these will still be done from my home.

Beckworth: All right. So George, I'm excited to get you on and talk about cryptocurrency today. It's been in the headlines. Of course, one of the big stories has been El Salvador and the legal tender law down there. We're going to come back to that, but I thought we would start by just going over some of the basics, and maybe help us think about cryptocurrency and Bitcoin in particular, since it's so popular. Let's start. I want to draw upon an article you wrote called “Synthetic Commodity Money” from the Journal of Financial Stability. We'll have a link in the show notes, so check out the article. This was really fascinating. But in this article, you classify different types of money. So maybe help us think through, what's fiat money, what's commodity money, and what's this synthetic commodity money definition you propose in the paper?

What is Synthetic Commodity Money?

Selgin: Well, David, first of all, I think I wrote the first draft of that in 2011, actually. Late 2011. At the time, I was struck by the fact that Bitcoin didn't seem to me to fit into any of the existing classifications of money. Essentially, there were two types of money that were allowed before the advent of Bitcoin and for a few years later, and those were fiat and commodity money. The distinction's pretty straight forward. They are on two dimensions. I recognize two dimensions. Fiat money is distinguished by having no non-monetary use value. There's not much you can do for it unless it's a medium of exchange, so it tends to become worthless once it's no longer accepted for that purpose. So no non-monetary use is one characteristic.

Selgin: The other one is that there's no natural scarcity. It doesn't take much to produce fiat money, and digitally it might cost practically nothing. So its scarcity, which is essential for it to have values since there's no non-monetary value, well, it wouldn't have value if it weren't scarce even if it had non-monetary uses, its scarcity has to be contrived. Someone has to see to it that the stuff doesn't get created in vast quantities, even though that's possible with little or no expense. On the other hand, you have commodity money. Commodity money. As the name suggests, obviously we're talking about gold and silver, but also wampum and cowrie shells and all kinds of things. Those commodity monies have, that is the commodities involved, have non-monetary uses. They were valuable as for things other than exchange before they became money. Once they cease to be money, they continue to have value for non-monetary purposes. So that's one thing that's true about them, different from fiat money.

Selgin: And then you have the fact that these commodities are naturally scarce. Some of them are reproducible, but not at zero cost. They tend to be reproducible at rising marginal cost. Some aren't reproducible at all, that is they exist only in some fixed amount on the planet and that's that. Okay. So those were the two categories. Bitcoin, it occurred to me, didn't fit for two reasons. On the one hand, like fiat money, it doesn't have any obvious non-monetary use. Here I should say that mere investment use, use as an investment vehicle, is still a non-monetary for the purpose of this breakdown. In other words, you're still holding it because somebody else values it in that case, otherwise you couldn't realize your profits from investment. On the other hand, its scarcity is in a sense not natural, but absolute. That is, nobody has to be put in charge of keeping it scarce, because by design, it is unalterably scarce.

Selgin: In Bitcoin's case, eventually there'll be 21 million coins or we will approach that limit asymptotically, and that will be that. Obviously, there's a real cost of production involved in producing Bitcoins, up to the point where you can't produce anymore and it's quite high. I mean, it's equilibrium to their value. So I call that category, that third category, synthetic commodity money. There's actually a fourth category of stuff that I also include in the table, because we have four quadrants really, and the fourth one would be something that has a use value like a commodity, but its scarcity has to be contrived. I think I give in the paper, it's been a while since I've read it, a Picasso print or some kind of artistic print where a certain number of copies are run off, it costs practically nothing to make more copies, but in order to give the prints value, the plates are destroyed, and then you only have so many examples outstanding.

Selgin: Anyway, now we have four possibilities, but the one I was interested in discussing in the paper particularly was Bitcoin. What I said about it, and what I still believe is true, is that it's not so much that Bitcoin was ‘it,’ that we discovered some secret new money or potential money that could solve all our problems. But what I argued was that what we have here in principle is the possibility of designing a synthetic commodity that combines the advantages of fiat money, or some of them, with the advantages of commodity money. It would combine them in this way. First, you don't have to worry about non monetary demand shocks. With gold, you have to worry about if there's a big jump in the demand for a gold, for industrial uses or for jewelry, the real demand, and you don't have to worry about supply shocks.

It's not so much that Bitcoin was ‘it,’ that we discovered some secret new money or potential money that could solve all our problems. But what I argued was that what we have here in principle is the possibility of designing a synthetic commodity that combines the advantages of fiat money, or some of them, with the advantages of commodity money.

Selgin: Ideally, you can do even better. You can design a supply algorithm that would be macroeconomically smart. So I was envisioning the possibility of something better than Bitcoin, and that it had those two advantages I just mentioned. But instead of the supply just growing asymptotically to a limit of 21 million, the supply algorithm, which is hardwired, has some feedback from some macroeconomic variables so that it could do something like maintain a stable price level, or maintain a constant value in terms of some commodity, or perhaps stabilize nominal GDP. That's the sort of thing I saw as a potential for this general category of money. So I wasn't extolling Bitcoin as the great solution to everything, but-

Beckworth: So synthetic commodity money is this ideal that you're aiming for in the paper? Would you even include Bitcoin under that?

Selgin: I would definitely call Bitcoin a potential synthetic commodity money.

Beckworth: Okay, potential.

Selgin: The reason I say potential is because it hasn't really been adopted widely as a medium of exchange. Its purpose is mostly investment, and it serves as a niche and medium of exchange. Whereas I go for the standard textbook definition of money as a generally accepted medium of exchange. Now, under the new law in El Salvador, which we're going to get around to talking about, it might qualify as money for that economy, for example. So in that case, it will be a synthetic commodity money, not just a potential one. But it's not a particularly well-designed synthetic commodity money. I should say, this is not meant to be a tough criticism, or I'm not picking on the designers. It's the first one, and it may turn out that it's very hard to design algorithmic cryptocurrencies that do what you want them to do, when what you want them to do is anything fancy. We know that algorithmic stable coins, for example, have been flops in a number of instances.

Beckworth: Yeah, I was going to ask you, are there any other cryptocurrencies out there that better fit your ideal of a synthetic commodity money?

Selgin: Yeah, that's a good question. Well, honestly, I don't want to say that there aren't, because it depends on what purpose. If it's my idea, let me put it this way, there may be synthetic commodity monies that work just the way they're supposed to. Yeah. I mentioned some stable coins that are supposed to keep stability in terms of the dollar have failed, but not all the algorithmic stable coins have failed. But my ideal would be one that keeps nominal income stable. The problem with anything that has to refer to some outside datum, like what's happening to nominal spending, is you either need an oracle, which is an external source of information that you're relying on, and then of course the trust in the oracle becomes a new level of trust that's needed.

Selgin: Some people don't like that. They want to have as little reliance on outsiders you have to trust as possible. But what you really need in this case is some way of having the cost of mining depend on the actual final transactions in Bitcoin that are being undertaken, so that if it becomes widely used, it would be a case where the algorithm is stabilizing total spending, total and final. That's the ideal. Long after I wrote the article, and for some time afterwards, I said, "If anyone can come up with a crypto that can do this-"

Beckworth: Supposed to be a nominal GDP coin.

Selgin: That's right. Let me know, let me know. By the way, I should have mentioned that part of the inspiration for my enthusiasm for cryptocurrencies, what they're doing or what Bitcoin and some others are trying to do is really just a high-tech version of a proposal Milton Friedman made back in the '80s. I think it was around '84 when he said, "Why don't we just get rid of the FOMC, replace it with a computer, and have the computer generate X percent growth every year."

Beckworth: That's an interesting way to think of it.

Selgin: In 1984, computers were only so-so. And then of course, the idea is you start the computer and then you throw away the controls and make it so nobody can go in there and do anything to stop the algorithm. So there's more to it than just having a smart computer. You'll have to make sure nobody has the, as it were, the keys to the computer. Anyway, I was thinking about that and I said, "Well, that's what Bitcoin has done." Bitcoin has done that. It said, "We've got a computer, it's going to spit out so many Bitcoins per year." Of course, it's not just a computer. It's a very complicated, but fascinating, impressive technology software, and nobody can change it. You can fork, you can have new versions of Bitcoin, but there's one that's going to keep doing what it was originally supposed to.

Beckworth: That's fascinating. So Friedman's K-percent rule in modern form is Bitcoin or some form of cryptocurrency?

Selgin: That's right, except with Bitcoin, it's a K-percent rule where K is steadily diminishing.

Beckworth: So just on that point, so as we get closer and closer to the 21 million, does the marginal cost of mining go up?

Selgin: Yes. The marginal cost of mining understood as the cost of producing another Bitcoin. It approaches infinity.

Beckworth: So it's like approaching the speed of light. The closer you get to it, the more dense you become, the harder it is to get to that final destination. So in theory, that 21 millionth Bitcoin will be very, very costly to produce.

Selgin: That's right. It would be unprofitable to.

Beckworth: Okay. So we'll always be in the limit with Bitcoin. So this also reminds me, you mentioned Milton Friedman. This reminds me of Friedrich Hayek's denationalization of money. Is there any overlap with his thinking on this, or is that different?

Selgin: I think there's a lot of overlap between Bitcoin and Hayek's view. Indeed a lot of people, if I may use this as an excuse to bring up one of my other hobby horses, free banking, a lot of people think Hayek was a proponent of free banking. Free banking is where the banks can issue any form of hand-to-hand or other media denominated in some underlying base money, which the banks themselves don't produce. You don't have one central bank that has a monopoly on any of these substitutes, redeemable substitutes. Hayek didn't like free banking. He actually was opposed to it. But what he did describe was a system where you have competition among fiat money suppliers. The competition, in his view, ought to include private fiat money issuers would try to compete with the official ones.

Selgin: Hayek believed, I think rather overly optimistically, that this rivalry would be a rivalry where the issuers who succeeded in achieving the most stable purchasing power for their monies would command the largest market. Well, private fiat money, understood strictly as fiat money produced with the old fashioned fiat money technology, not synthetic commodity money, fiat money has a lot of issues. Larry has been one of several people, Larry White, pardon me, who has explained that it's quite possible that the profit maximizing amount of fiat money that a competitive supplier would produce could be highly inflationary, even hyperinflationary amount. Basically, get suckers to take your money on the promise that you're not going to let it depreciate and then depreciate it like mad, and then you're getting all these assets for practically nothing.

Selgin: Anyway, however, that may be why we have never seen private fiat monies. But it turns out that cryptocurrency, and Bitcoin in particular which Hayek could not have anticipated, is the closest thing, in fact, to Hayek's hypothetical idea of a private fiat money, except it's more clever and provides us more strict guarantee of limited supply and scarcity. Therefore, it avoids the problem of confiscatory over-issue. So it overcomes. It's able, actually, to thrive, and Bitcoin has shown that. So in a way, Hayek was anticipating Bitcoin when he talked about private fiat monies. But what we know now is that no, it's not clear that private fiat money is viable, but private synthetic commodity money is viable, where viable means it can have a positive value in equilibrium.

Beckworth: We'll probably link to that book, The Denationalization of Money. One last question on this before we move on to El Salvador. So you call this category synthetic commodity money, but going back to Hayek, why not call it synthetic fiat money or some other term? Why is commodity in the name?

Selgin: I want to, again, emphasize the distinction between the kind of stuff that Bitcoin is and fiat money, where fiat money doesn't have an unalterable scarcity. That's the thing. With fiat money, we need a name for the stuff that's fiat money, where someone has to be in charge who we trust to put limits on the supply, where the profit motive alone won't cut it. That's the thing. We need a name for that money, and we've got one. It's fiat money. So I wanted a name, a different name, to give to these things where the scarcity is like the scarcity of a commodity, in that nobody is in a position to just make a decision to change it. Nobody's in a position to say, "Let's have more gold in the ground than is presently there. Let's double the supply of gold."

Selgin: I don't mean the mine supply, I mean the total supply, and therefore the cost of mining gold we're going to lower. Nobody can do that. With fiat, you can. Well, nobody can change the cost of mining Bitcoin. So it's like gold in that respect. So I think we need the name fiat money to describe what we've been calling fiat money all along, but if we call Bitcoin a synthetic fiat money, I think we obliterate an important distinction between it and fiat money, the distinction Bitcoiners themselves are extremely keen to emphasize, which is you don't have to trust anybody to keep it scarce.

Beckworth: So let me ask the question this way, George. Bitcoin has both some elements of fiat and some elements of commodity money. You chose synthetic commodity money. Maybe one could say it has more commodity money characteristics than fiat money, so therefore it makes sense to call it a synthetic commodity money?

Selgin: I'm not sure I would put it that way, David. I think the reason why I like synthetic commodity money is because it suggests that it's as if you went into the laboratory and came up with your own, using test tubes, you invented a new scarce thing that never before-

Beckworth: Alchemists.

Selgin: Like alchemy, yeah. I once put it, I think in that paper, I said, "What if you were God, and you've created silver, you've created gold, you've created wampum, and now you want to create something that is going to have the perfect supply behavior so that people discover it only when it helps to stabilize nominal GDP?" Because God, as everybody should know, is an NGDP proponent. Has he gotten his mug yet, David?

Beckworth: I think he owns all the mugs already, so he's covered.

Selgin: Okay. Anyway, so what we have here is the possibility of coming up with brand new commodities that behave in different ways. There is a sense in which they're not commodities and are like fiat, because they don't have any non-monetary use. In that sense though, I think that Bitcoin is more like a commodity. Okay. Also, we think about the whole investment thing with Bitcoin, it sure looks more like a commodity, the hedge potential and all that, than a fiat. Indeed, well, I'll just say that I like the term better than I like synthetic fiat. Fiat is already a fake synthetic thing. That's the other thing. Fiat is a human creation, so it's synthetic in some sense already.

Beckworth: Okay. Just a very basic question about Bitcoin. How is it being used right now? Isn't it more of an investment asset?

Selgin: Mostly.

Beckworth: Mostly, okay.

Selgin: And a niche medium of exchange for certain kinds of activities.

Beckworth: This is a nice segue. This could all change because of El Salvador, at least potentially. So walk us through what happened in El Salvador. It's going to change the world, it's going to finally create the network effects for Bitcoins. Tell us what's happened down there.

What Happened in El Salvador with Bitcoin

Selgin: Let me step back to do that, David, by pointing out that I have been called a Bitcoin skeptic with justice, not because like some people I go around saying, "Bitcoin is worthless. It's going to go to zero. It's a scam. It's a Ponzi scheme…" I've never been keen on those arguments. What I have said is that it lacks the potential, in my opinion, of course. I'm speculating when I say this, lacks the potential to become widely adopted as a medium of exchange for ordinary purchases, to go buy your coffee, to go to this, to pay for your bills and all that. And as I've tried to emphasize, the fact that it has appreciated a lot in value since its introduction doesn't itself make it attractive for those general purposes, and doesn't more specifically mean that it can overcome the huge advantage an incumbent, well-established money generally has, because of its established network.

Selgin: So, network economies are very important here, and of all the monies in the world, the dollar has by far the biggest network worldwide. So it's the hardest currency for any alternative upstart to unseat or compete with. Okay. Now, sure enough, El Salvador has tried to throw a wrench into my argument, but not really. Because what they've done in El Salvador, what the government has done, is to pass a law that not only makes Bitcoin legal tender in the standard sense of the term in El Salvador, that is, something that can be used to settle outstanding debts, but also makes it a compulsory tender. That is, anybody who is equipped to receive Bitcoin in exchange for goods that he or she is selling, is supposed to accept it, is legally required to accept it. That's Article 7 of the Bitcoin law.

The fact that it has appreciated a lot in value since its introduction doesn't itself make it attractive for those general purposes, and doesn't more specifically mean that it can overcome the huge advantage an incumbent, well-established money generally has, because of its established network. So, network economies are very important here.

Selgin: So what the government is doing, and particularly what the president Bukele has done in El Salvador, is to jumpstart a Bitcoin standard there, using the law to overcome the network disadvantages that would make it practically impossible that Bitcoin would have spontaneously, yes, displaced the US dollar, which had been El Salvador's official currency since 2001. Let me be the first to say, in case others reading my criticisms of what's happened in Bitcoin, which I know we're going to get around to, but unless they should think that I find this decision of Bukele's entirely incomprehensible. I want to make clear that if I were he, and my goal was to give Bitcoin a chance of becoming El Salvador's money, I might have done just about all the things he's done.

Selgin: I think he's rushed it. I think we're all seeing evidence there are bugs in the arrangement that has been set up, and they're starting to get those ironed out. But generally speaking, the strategy has been absolutely good, given that end. So forcing people, merchants to accept Bitcoin if they're well equipped, seeing to it that more of them get equipped, providing the wallets that are necessary, public wallets. Not necessary, but that are alternatives to other wallets that can also be used in the system, and removing the risk of accepting Bitcoin from the merchants and placing it on the broader public. That's also a very useful device. Buying and distributing ATM machines around the country. All of these things are exactly what should be done if you're going to get Bitcoin used. Now, that doesn't mean I like what's been done there or that I think it's good for the Salvadorian people. I like some of what's been done, but I have real problems with the coercive aspects of the law.

Beckworth: Now, part of what they did is they actually gave out Bitcoins to every citizen. Is that right?

Selgin: Well, no, not quite. What they did, there are people reporting that everybody in El Salvador is going to get $30 worth of Bitcoin. It's not true. The deal is, if you download the government's wallet, which is the Chivo wallet, then you are given an initial balance of $30 worth of Bitcoin, which you can spend on goods or services, or you can convert them to US dollars also on your app. So they're actually dollar stable coins, and then you can go an ATM or whatever. But that's only if you download the Chivo app. So far, the downloads are reported to have exceeded 500,000, half a million, but the population of El Salvador's is much larger than that. 500,000 is about six or seven percent of the total population.

Beckworth: For listeners who don't know, why is it important to have this digital wallet when you have Bitcoins?

Selgin: So in order to use Bitcoins, both the merchants and the buyers have to have some digital wallet to store the Bitcoin, as it were, we're talking about storing digital media. Now, these wallets aren't devices. They're just apps. You put them on your phone. That way, the information can pass from the wallet in one phone to the wallet in another phone, and that the credits get moved. We're really talking about, these Bitcoin of course exists only in the wallets and the phone. If people convert to dollars, they have essentially the equivalent of dollar credits with the government. What's interesting is my immediate reaction to this law, to Article 7 in particular was, well, this is terrible because there's so much risk involved. In fact, since the law was implemented back in June, there have been huge gyrations in Bitcoin's price.

What they've done in El Salvador, what the government has done, is to pass a law that not only makes Bitcoin legal tender in the standard sense of the term in El Salvador, that is, something that can be used to settle outstanding debts, but also makes it a compulsory tender. That is, anybody who is equipped to receive Bitcoin in exchange for goods that he or she is selling, is supposed to accept it, is legally required to accept it.

Selgin: Merchants had been receiving Bitcoin in their exchanges who didn't really want Bitcoin, and polls have shown that most would have just as soon stayed with dollars, not surprisingly. Then by the time they converted into dollars, apart from fees that would have been involved in the conversion, even a short interval between the sale and the conversion could have meant taking some big losses. Now, the government has addressed that in the actual Bitcoin law, but it has done so, it has assumed the risk, but let's be clear, governments don't actually assume risk. What they do is they transfer the risk to other parties. So in this case, the government has set up an exchange fund and an exchange bank, as it were, and it will instantly convert merchants' Bitcoin receipts into dollars at the original dollar posted price, because the dollar is still the unit of account in El Salvador. So, prices are posted in dollars.

Selgin: So the merchants get their dollars, at least they get virtual dollars. Then, and there is some risk there, because these virtual dollars are not necessarily fully back. They're essentially claims against the El Salvadorian government, so it's not as if it's no risk at all that they could turn out not to be fully backed, or that there could potentially be a convertibility problem. But in any event, the government has its own trust fund, originally funded with $150 million worth of Bitcoin. So it turns around, well, it can hold the Bitcoin that it has received, but it is assuming the risk. It's the one that's actually investing in the Bitcoin if the merchants choose to have dollars.

Beckworth: This is your point that the risk of this transformation to Bitcoin usage is being borne by the public via the government.

Selgin: It is ultimately being borne by the public. So the public has paid, first of all, for the initial $150 million investment in the fund, and the public will pay implicitly for any losses incurred on the government's Bitcoin holdings. Obviously, if the value of Bitcoin goes up, then the public gains to that extent, other things equal. But here's the hitch. The hitch is that when Bitcoin is expected to appreciate, then more merchants can be expected to just keep it in their wallets. When it's expected to go down, as when it's actually falling in value on the market and they have a big sell off, that's when they're going to be offering it to the government. So the government's going to be, as it were, buying in the down markets and not buying in the up markets.

Beckworth: It's procyclical.

The hitch is that when Bitcoin is expected to appreciate, then more merchants can be expected to just keep it in their wallets. When it's expected to go down...that's when they're going to be offering it to the government. So the government's going to be, as it were, buying in the down markets and not buying in the up markets.

Selgin: Yes. I think that what that means is that there's a considerable risk that the trust fund will be depleted, and it will have to be replenished, and then the taxpayers will take another hit. So I think it's this dangerous scheme, but of course it should be obvious, right? If something is risky and the government is covering the risk out of using public funds, that means there's no free lunch. Somebody is likely to lose money.

Beckworth:  Well, if it does become widely used as a medium of exchange, it becomes money in El Salvador, that would, I think at least mitigate some of that risk, right? Because the demand for Bitcoin would be up, and that would increase the value.

Selgin: It'd be a very small part of the whole world, because the global market for Bitcoin is much bigger. We saw this, actually. On the day that the Bitcoin law took effect, on the 7th, this is kind of a funny sad story, but there was this great effort on the part of the Bitcoin community to-

Beckworth: Get the value up?

Selgin: ... Get the value of Bitcoin up. Everybody's saying, "Let's express our approval and support for what's happening in El Salvador by all going out and buying Bitcoin. Let's all buy some."

Beckworth: So, like GameStop.

Selgin: Yeah. Sure enough, the value of Bitcoin collapsed 17% that day, at least for part of the day. So that's an illustration of the fact that El Salvador's a small part of the Bitcoin story in the world.

I think that what that means is that there's a considerable risk that the trust fund will be depleted, and it will have to be replenished, and then the taxpayers will take another hit.

Beckworth: This sounds very familiar to the international gold standard, right? A country that has gold as its reserve, as its unit of account, is going to have challenges if a global demand for gold varies or it goes up or go down, like the interwar period.

Selgin: That's right. Now, it's true though, but in the case of the gold standard, what made it relatively stable or one of the things that made it relatively stable, was the fact that it was many countries. Now, if many countries are on the same standard, that does tend to stabilize its value. But if one small country is, that doesn't. Look at the argument against dollarization, which is what El Salvador relied upon, until now, for a couple of decades. The argument is that if you're a small country and you're on this dollar standard, you just have to take what's given, all these fluctuations. The gold standard, same thing. If El Salvador were to adopt gold as its monetary standard, that wouldn't make the gold any less volatile a standard than it is today, would not return us to the stability of gold circa 1890. So one country going on a Bitcoin standard is like one country today going on a gold standard.

Beckworth: Well, I'm sure all the Bitcoin bands would say, "Yes, but El Salvador will be the first. There will be many more to follow."

Selgin: That's right, yes. Well, there, I would say too, that just as El Salvador is small, relative, its demand for Bitcoin is a small part of the total world demand. It's not that capable of driving the market. The same is true for the network effects, that yes, now you have a bigger network of people willing to receive Bitcoin as a medium of payments, but not enough to convince the rest of the world that this is now the network worth joining.

Beckworth: No, but it might send a signal that hey, they've tried it, let's try it too. Not that you're tapping into a bigger network, per se, but that you're part of the fad.

Selgin: Yeah. Well, if you were going to do that, I think it would be wise to wait a year.

Beckworth: Well, no. No doubt. I'm saying what the Bitcoin would say.

Selgin: They're already saying that. They're already saying that Argentina and a bunch of other countries are going to jump on that.

Beckworth: So all of Latin America will one day be on a Bitcoin standard.

Selgin: Yeah, yeah.

Beckworth: Let's move on, George. Let's talk about something related to this, and that's stable coins and Fintechs. You mentioned stable coins already, and Larry White was on the show previously talking about them. But the stable coins is a former cryptocurrency, tries to stabilize the value of the cryptocurrency relative to some underlying asset, often the dollar. Fintechs are doing something similar. You had a proposal for how they could operate, and it has to do with having access to the Fed's balance sheet. So walk us through that proposal.

Dollar Substitutes, Stable Coins, and Fintech

Selgin: Okay, yeah. Well, the proposal is not entirely related to stable coins, but more generally to a dollar substitutes. They don't have to be token based. They could be more like conventional bank deposits. Indeed, in my specific proposal, they are conventional bank deposits if you allow that a special purpose bank is a conventional bank for the purpose of that discussion. What I've argued is that, well, let me put it this way. We have a setup right now where banks, that is just plain old banks, not special purpose banks, are the main suppliers of media of exchange substitutes for Federal Reserve dollars, base dollars. Then you have stable coins that are operating outside of the Federal Reserve system quite independently, and outside of the general umbrella of regulations and all that. This is what makes people worry about them a lot. I should add, we also have money market funds as well-

Beckworth: That do something similar.

Selgin: That do something similar, but are subject to some regulation. I think it's very important to understand why the stable coin issuers exist, why there's a market that's been growing very rapidly for stable coins, and the reason is that they can serve some purposes that ordinary banks don't serve very well. Particularly for exchanges involving cryptocurrencies, where ordinary banks don't want to get involved with that. They're worried about the regulatory backlash, et cetera. So you don't have alternatives to stable coins that can do all the things that the stable coins are able to do. Now, some of that is also people trying to avoid know your customer or anti money laundering restrictions and that sort of thing.

I think it's very important to understand why the stable coin issuers exist, why there's a market that's been growing very rapidly for stable coins, and the reason is that they can serve some purposes that ordinary banks don't serve very well. Particularly for exchanges involving cryptocurrencies, where ordinary banks don't want to get involved.

Selgin: On the other hand, we have the movement for having central banks issue their own digital currencies for retail use. What we don't have, and what we need, are more opportunities for private sector, monetary or payments service innovators that are not conventional banks, to offer digital retail exchange media that can do things that ordinary bank deposits can't do, or serve markets that those bank deposits can't serve, or what the banks won't serve, and that can do so in a highly efficient way. This includes being able to supply mobile digital monies and that sort of thing. We've seen how effectively that's been done in many other countries. We should be able to easily go around and pay for stuff using our telephones and without having to resort to stable coins or to Bitcoin instead of dollars, anything like that.

Selgin: Okay. Well, there's a middle ground solution that's neither relying totally on the Fed to supply retail, digital, person to person payments media, and it's not relying on stable coins that have nothing to do with the established legacy US dollar system. That is: let fintech firms have accounts with the Fed, even though they're not fully fledged, conventional commercial banks. Let them supply payments media where the Federal Reserve is supplying the wholesale back room settlement system, and the money can move around through the Fed settlement system, but the retail media are provided by these private sector firms that really can do it very efficiently, so the Fed doesn't have to learn the retail digital money business, which is a big learning curve. To do that, you need to give these fintechs access to the Fed's books. You need to let them have whole master accounts at the Fed.

Selgin: So what I'm proposing is a compromise that could be summed up this way. Let's have central bank digital currency for retail, fintech suppliers of retail digital currency. That's one way you could think of it. Let them have the front end, let the Fed cover the back end. That's what firms are trying to do, some fintechs have been trying to do by acquiring special purpose bank charters, either from states, especially Wyoming, but also some other states that offer these charters. Or from the OCC, which has a couple of different kinds of special purpose charters. And then once they have these charters, they can apply for Fed Master Accounts, that is, to keep their money at the Fed like banks do and certain other agencies. And then they can settle payments on the Fed's wholesale rails through the Fed wire, particularly. In this way, we let these firms have the best opportunity to use their technological know-how to reach a broader public, and provide services that banks don't supply very well, including cheap payment services, including payment services for cryptocurrency, exchange, activities and mobile money, et cetera, et cetera, et cetera.

Selgin: The only reason this is controversial, and it's not a small reason, is the banking lobby particularly are saying we shouldn't let anyone have a Master Account and have dollar deposits or their equivalent unless they are subject to the same regulation as banks. But that doesn't make any sense if the holders of these Master Accounts, if the special purpose banks are not engaging in the same risky activities as ordinary banks, then when you come right down to it, there's one major risk ordinary banks take that's driving all the special regulations and that's maturity transformation. It's taking short duration liabilities, deposits, and using them to fund longer term loans or investments. There's always risk involved in that activity. So it's mainly for that reason that we have deposit insurance and strict capital requirements and so on.

There's a middle ground solution that's neither relying totally on the Fed to supply retail, digital, person to person payments media, and it's not relying on stable coins that have nothing to do with the established legacy US dollar system. That is: let fintech firms have accounts with the Fed.

Selgin: What I propose is that the Fed grant licenses pretty much automatically to any special purpose bank that commits to having its deposits 100% backed by Fed balances, by Master Account balances, so there's no maturity transformation going on with those deposits. In that case, you still want the banks to have some capital, because there are costs to winding up at institution for any reason. I suppose it fails because of malfeasance or who knows what, right? So you want to have capital requirements, but you don't have to have quite as strict capital requirements. None of these special purpose charters allow the banks set up under them to have no capital. In fact, they have pretty substantial capital requirements. Deposit insurance is completely unnecessary when you have no maturity transformation risk.

Beckworth: There's no risk, because the Fed's backing the account.

Selgin: Your money is in the Fed. Yeah.

Beckworth: So what would be the appeal of these accounts? They're very safe. I mean, why would I as a retail customer go there?

Selgin: You wouldn't go just for safety. Safety is essential. So to attract you and to make you use them instead of ordinary bank accounts, safety is important, I should say, but you're going to do business with these banks not just for that, but because they're providing special payment services that other banks don't. Particularly you can do crypto exchanges, et cetera. Some of the things that these special purpose banks would do, other fintechs could do without acquiring banking charters if they could get a commercial bank to act as an agent for them. So that's the other possibility, but it's more expensive, and fewer and fewer banks want to deal in certain kinds of-

Beckworth: If you've got to have a middleman to give you access to the Fed's Master Accounts, it's more expensive. That's business.

What I propose is that the Fed grant licenses pretty much automatically to any special purpose bank that commits to having its deposits 100% backed by Fed balances, by Master Account balances, so there's no maturity transformation going on with those deposits.

Selgin: There are also fewer and fewer ordinary banks that want to get involved in those things, because they fear the regulators.

Beckworth: So I could have an app on my phone in the future that gives me access to the Fed's balance sheet?

Selgin: That's right. You'd be dealing with some private payment service supplier that is not an ordinary bank. It might be a special purpose bank. It presumably would be. Under current law, the Fed cannot give a Master Account to a non-bank or non-depository institution, so you have to have a charter that says you're a bank, whether you're an ordinary bank or a special purpose bank. So I'm taking that for granted. But you could have an app on your phone from a special purpose bank, which makes no loans. If it's taking deposits, it makes no loans.

Selgin: Now, some special purpose banks might make loans, but they can't fund them with deposits. That's a different category. That would still be okay. But in any event, it doesn't make loans. Everything is 100% invested. What this fintech does for you, the special purpose bank does for you, is that it helps you to undertake exchanges in the cryptocurrency markets and another places where ordinary banks are not willing to take you on to provide that service. Or it's providing you access to real time payments. There's some advantages involved in the payment services that it's doing. It's the fact that it's on your phone. It's the fact that it's real time. It's the fact that you can buy crypto. It's something that ordinary banks don't do very well at all.

Beckworth: This is interesting for a number of reasons. One, there's been people like our friend, Morgan Ricks, who's wanted a Fed account. This would be one way to provide it, but it would also be one way to do it without some of the concerns that people have with central bank digital currency. So I've been thinking about this, talking to people about central bank digital currency. What are some of the issues we think about, and I'm sure the Fed itself is now thinking about it, because it has this paper coming out shortly, a study paper on central bank digital currencies. So here's just a brief list I came up with why we might be worried about the Fed providing retail central bank digital currencies. One would be privacy issues. Maybe another would be, is this the backdoor for the Fed to do negative interest rate policy? Banks might worry about losing business. Technological innovation, would the Fed be able to continue to innovate?

Beckworth: And then the cost issue. The Fed by law has to recover cost, and what you're proposing I think at least addresses some of them. It definitely addresses concerns about technological innovation. It would cover the cost issue. Probably would be fine with the privacy issue, and maybe banks might still have some objections, but there would still be a private sector, so that the private financial intermediation concern would go away. Maybe you wouldn't address the negative interest rate issue, but you would, I think, address a lot of the concerns that critics have with this approach. Is that right?

Selgin: I think this would address a lot of the criticisms, if the critics would be willing to recognize the fact. Actually, part of this got started when my friends at the Bank Policy Institute criticized proposals of the application by Kraken Bank, which is a Wyoming SPDI, for a Master Account at the Fed. They said, "they shouldn't grant the Master Account to Kraken because Kraken is going to be risky. Wyoming's SPDI law doesn't require them to hold as much capital as banks and blah blah blah. If it quacks like a duck and walks like a duck, et cetera." So they're arguing that Kraken wants to be a bank without having to play by the rules. Well, in fact, I went back to these guys. They specifically said, I should add, that the problem is maturity transformation. That's a bank. I said okay, well, suppose that Kraken says, "We're going to do 100% Fed balances. We're not going to do maturity transformation." Now what's wrong with giving it a Master Account?

Selgin: I got a kind of Ralph Kramden response. Anyway, it turns out that in fact, Wyoming's law, it's been modified a little bit, but Kraken could never have made a loan. Under current Wyoming law, it could only invest in high-quality liquid assets. But in fact, they'd be willing to hold 100% Fed balances. Their business model doesn't require them to. There's no need for them to hold anything else. First of all, because for one reason, it's that reserves pay just as much as other high-quality liquid assets. Sometimes they pay more. They seldom pay a lot less. So why should they care? As I was saying before, their customers aren't doing business with them because they want a high return. These are payment services. It's transferring funds. They're not there to earn a return on their balance, so it's not important for most of these SPDIs to be able to invest in commercial paper, let alone make loans. They don't need the return. That's not their business model.

Beckworth: This is, again, fascinating, because it does address many of the calls for a central bank digital currency. At the same time, it addresses the concerns. For most, I mean, there'd be some groups, as you mentioned, that may still object, but it's fascinating to think that the people who want Fed accounts might be excited about your idea, while the banking industry may still have some reservations about it. But this may be the modest step forward. The Fed's not going to do something really radical. If I had to make a guess about what they're going to do with CBDC, it's going to be a modest step, because they're a conservative, small seat. They want to be careful. They don't want to rock the boat. This could be the middle ground that they land on.

Selgin: Yeah, I think it could be, and I hope they'll give it serious consideration. I think that the Fed is very poorly equipped to offer directly retail digital payment services. Even many of the proponents of central bank digital currencies would want them to be supplied in directly through the banking system. This is hardly any different, except why limit it to banks? It's important not to limit it to ordinary banks, because we have to recognize there's a lot of different markets digital currencies conserve. We need specialized firms to take full advantage of it and do it differently, different ones offering different services, and we want them to keep on innovating. We want to make sure we don't exploit technologies that allow more things to be done with digital currency than we know can be done now. All those things require having a private competitive system. It's heterogeneous.

Selgin: The Fed is one firm. It might have different branches, but it's not obvious at all that it could, I can't imagine it providing different kinds of specialized digital payment services. There's going to be one business model that it's implementing, whether it deals with banks or not. I don't think we need that. I think we need new a plurality of digital payment service suppliers. I think if we had that, there'd be little, very little if anything, that the Fed's own product would be capable of adding to the mix. This doesn't mean that you might have some goals that you can only achieve through public policy. For example, it may be that you want some of these products to be accessible to people who can't afford them from the private market, given the cost structure. So you want to subsidize it. Okay. Subsidize it. But that's different from having the Fed be the supplier.

Beckworth: Well, George, this has been a fascinating conversation, and unfortunately, we did not get to your piece on discussing the higher inflation targets. We'll provide a link to it, and we'll have you come back on the show though next episode, number 10 for you, to discuss why you are hesitant for the Fed to adopt a higher inflation target. But that will be another fun conversation. But today. We covered a lot of ground. Bitcoin cryptocurrency, CBDC, stable coins, and I think this will be an area we'll come back to visit again in the future. So thank you for joining us today, George.

Selgin: It's been a lot of fun, David. Thank you.

About Macro Musings

Hosted by Senior Research Fellow David Beckworth, the Macro Musings podcast pulls back the curtain on the important macroeconomic issues of the past, present, and future.