May 24, 2021

George Selgin on Inflation, Fintechs, and Broadening Access to Fed Master Accounts

Granting fintechs access to Fed master accounts could be an important step toward creating a financial landscape with more innovation, better services, and lower costs.
David Beckworth Senior Research Fellow , George Selgin Senior Affiliated Scholar

Hosted by David Beckworth of the Mercatus Center, Macro Musings is a new podcast which pulls back the curtain on the important macroeconomic issues of the past, present, and future.

George Selgin is the director of the Cato Institute’s Center for Monetary and Financial Alternatives and is a returning guest to the podcast. He rejoins Macro Musings to talk about the Fed’s recent calls for comments on opening up Fed accounts to fintechs and other non-bank financial firms. George and David also discuss monetary plumbing issues, the state of fiscal QE, and 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 macromusings@mercatus.gmu.edu.

David Beckworth: George, welcome back to the show.

George Selgin: Hey, David, it's really great to be back.

Beckworth: It's great to have you on. I believe you are now the seventh time on the show, setting the record for the number of guests and it's been great to have you on here. I'm excited, George, to have you on because among other things we just learned that the CDC will now allow people into buildings without masks if they are vaccinated, which means, George, you and I get to go to conferences again, right? We get to attend presumably the Cato Monetary Policy Conference again and if the Kansas City Fed folks are listening, we would love to attend the Jackson Hole conference again. So, this is an exciting time. We might be getting a little bit closer to normal.

Selgin: Yeah, I'm hoping the monetary conference is live this fall and I'm looking forward to doing my next seven Macro Musings live.

Beckworth: Yeah.

Selgin: And maybe [inaudible] in the next few weeks.

Beckworth: Right, right. I'll let our guests know that George is finishing a book on the New Deal, Great Depression era and we'll make that entire show at some point in the future. But I'm glad to get you back on, George, because I really wanted to talk to someone about the Fed’s announcement about allowing non-bank financial firms to have access to Fed master accounts, which is a pretty big deal. It has allowed, to some extent, some other non-bank financial firms access to the balance sheet but this would be a big step forward because it'll allow a lot of fintechs and other entities to do so.

Beckworth: Now, before we get into that and that's going to be the heart of our discussion today, I also wanted to speak to another recent development. So, we're recording this show on May 13th, on a Thursday, but just yesterday, George, as you know, the CPI inflation numbers came out for the last month and they were quite a bit higher than expected. So, there's stories we can tell, explanations we can tell, base effects, it’s kind of catching up from the pandemic. There's supply bottlenecks. Lumber, oil. A lot of things we can attribute to this kind of one off developments that may be with us for a few months but it also could be in part due to the Fed’s new framework, right? The Fed actually calls for some temporary inflation overshoot given its average inflation targeting framework, which I've been calling all over Twitter really a watered down version of price level targeting and I'm curious to hear what you think these recent developments mean for the Fed. What do you see happening and where do you see it going?

Implications of the Recent Inflationary Spike and the Fed’s New Regime

Selgin: Well, there are three schools of thought as I see it, David, perhaps there are more but there are at least three concerning what the recent inflation number means. One school says, "Haha, told you the Fed had overdone it last year. Now they're losing control of inflation and we're in big trouble because it's going to keep climbing." The second school says, "No, no, no. The numbers were a little higher than usual but they are still consistent with the Fed’s inflation objectives. There's some overshooting that's short run but you need that when you're doing average inflation targeting so there's really nothing to worry about. Everything's fine."

Selgin: The third school, which is the one I belong to says, "Well, the Fed is still on target. That is, it hasn't over eased yet. But there is something to worry about." And the reason I think there's something to worry about, as I said, it's not because the Fed has already blown it but I'm worried that because the Fed has announced in the past that it didn't plan to raise interest rates for some time and has announced even recently that it doesn't plan to do any tapering that it may put itself into a corner if rates rise ... Sorry, if the inflation rate rises faster and continues to do so. It may need to go back on some of these utterances to stay within its long run inflation target and I'm worried about there being some risk that it won't tighten when it needs to because it has provided forward guidance that suggests that that tightening won't happen and that's a worry of mine.

I'm worried that because the Fed has announced in the past that it didn't plan to raise interest rates for some time and has announced even recently that it doesn't plan to do any tapering that it may put itself into a corner if rates rise ... Sorry, if the inflation rate rises faster and continues to do so. It may need to go back on some of these utterances to stay within its long run inflation target and I'm worried about there being some risk that it won't tighten when it needs to because it has provided forward guidance that suggests that that tightening won't happen and that's a worry of mine.

Share this

Selgin: There's more to it. I think the fiscal situation is also a great concern because there too you have some pressure on the Fed to not tighten when it should. I don't know how strong that pressure is but it's a concern whenever the debt is so large and the interest burden is going to be so sensitive to any upward interest rate adjustments.

Beckworth: So, just to summarize. Your concern is that even though the Fed has said, "We will tighten policy once inflation is running above two percent for a sustained period and once we get to what we think is full employment." So you're saying, even if we get to those milestones and get the data that the Fed may fail to tighten policy because it's committed to some kind of calendar based guidance. Is that the concern?

Selgin: Yes, it's a little more complicated than that. Part of the concern is that yes, they're not exactly committed to calendar based guidance. They have tried to have it both ways. They've said that new policies are data driven in the sense that they're only going to hike rates when they see inflation go up for some length of time persistently. But on the other hand, they have said that they don't expect to raise rates for some time even 2023 in some earlier pronouncements and that's a kind of forward guidance that is, if not inconsistent with a data driven policy, is certainly is going to raise some tensions. That is the Fed could say, "Well, we didn't absolutely promise we wouldn't raise rates earlier," but they're still going to be guilty, I think, of giving the public the reasons to expect that they won't tighten even when they probably should. So, that's one factor. But there's another factor, which is just as important. The Fed's average inflation targeting framework is a very vague framework. It's not just vague because it's average inflation targeting rather than price level targeting. It's vague because they don't say how far back they're going to look in deciding how far they've got to go to get the price level back up where they want it.

Selgin: They also haven't said how quickly they'll get the price level back up once they decide to do so. What that means in practice is that we can't tell whether the Fed is doing what it said it would do in response to the data. That is, we can't tell for awhile, at least, and perhaps for a good long while. The Fed could in fact, let's suppose it knew what the parameters are, let's suppose they know and they probably don't. Let's suppose they know the parameters, they're just not telling us. Then we can't tell whether they are actually not responding to the data the way their own parameters tell them they should or they are responding but that gives people reason to worry even if the Fed is doing what it really planned to do all along. So, I think the whole setup is shaky and I think it could unnerve people and make them form expectations or the opinion that the Fed is not really committed to targeting inflation even if it actually is and that will just make it harder for the Fed to successfully target inflation.

I think the whole setup is shaky and I think it could unnerve people and make them form expectations or the opinion that the Fed is not really committed to targeting inflation even if it actually is and that will just make it harder for the Fed to successfully target inflation.

Share this

Beckworth: Yeah, I'm sympathetic to that point there. Clarity might go a long ways in getting us to a fully credible average inflation target. But I also understand maybe why they've been reluctant to add that clarity. And that is, maybe it's impossible to do so. Maybe on a conditional basis you could put out some numbers but I had this conversation with Bennett McCallum that reminds me of what we're talking about now. As you know, George, Bennett McCallum was a big advocate of nominal GDP targeting as we are and he was very generous with his time toward me and would tolerate my emails and questions years ago. And we got in these discussions of whether we should do nominal GDP level targeting or just regular nominal GDP growth rate targeting and he didn't like level targeting and I did and the reason he didn't like it as best I understood was that it allows for a lot of discretion.

Beckworth: During that catch up period, there's a lot of wiggle room, a lot of space for uncertainty, for things to go sideways and I think that was one of the big reasons he preferred a growth rate target. It's a little more transparent, a little clearer, easier to understand. You don't have all these parameters you've got to define. So, I'm sympathetic to the point you're making. And I also think that this is a good reminder that the Fed has done this same error but in the other direction. You wrote about this, right? The 2015 rate hike. You mentioned in your paper about Janet Yellen and the rate hike. That they did that just because they felt they had to, because they had to normalize monetary policy even if the economics wasn't behind that. So, is this kind of like a flip side of that mistake that you're worried about?

Selgin: Yeah. I think it is. I think the problem is there's a conflict between having a strictly data driven policy and providing forward guidance, particularly unconditional [guidance]. We don't plan or we don't foresee rate increases for some time, for a year, for two years, until 2023, whatever. Any statement like that is a problem when you're trying to implement a strictly data driven policy. You can't have it both ways. But the other thing I would add is that if it is perhaps true, as Bennett McCallum says, that any kind of level targeting is going to pose problems of ambiguity, but of course, it doesn't have to be as ambiguous as Fed’s present policy is.

I think the problem is there's a conflict between having a strictly data driven policy and providing forward guidance, particularly unconditional [guidance].

Share this

Selgin: The Fed could say, "Look, we're going to go… look the inflation rate has been at two percent or turns out to be two percent or more looking back x months,” where x is a number. It could be 12, it could be 24, it could be something bigger. “Then we're going to start hiking." That's concrete and that allows you to say whether they're doing what they said they would do or whether something's going on that makes them reluctant to do what they said they would do, something that perhaps we should be worried about them taking into account in their decisions. So, it could be a lot better than it is. I don't really know. I'm sure there are theoretical arguments for wanting to be flexible about how far back one looks but I suspect they're not so strong as to make it and make committing to a window size the best policy. So, I don't know what their arguments are for not doing that but I think it's a mistake.

Beckworth: A couple of comments on that. One, I recall Governor Lael Brainard saying in an interview when she was asked about this process that the review they just went through and the new framework. She mentioned that in five years it might be good to have another review to look back and see what this new framework has accomplished. So, that at least, gives me at least a ballpark of how long they… at least Governor Brainard, how long she sees this taking effect. In other words, she hopes to see some traction, some measure of success within those five years but more clarity definitely would help.

Beckworth: Let me push back, George, just a little bit on your statement about their forward guidance because I know there will be some listeners out there who might be thinking this and maybe this is something that only Fed watchers, like ourselves and others, who closely watch the Fed are aware of. But the Fed’s summary of economic projects where you get a lot of this information about their forecasts for inflation, for interest rates. Those forecasts are conditional aren't they? They're conditional upon the current economic environment. If that's the case, is your critique though that they're being interpreted differently? They're being interpreted as an unconditional forecast?

Selgin: My critique is that they lend themselves to being interpreted so and they have been interpreted so. I've talked to several people in the press. Just today I had a conversation with a very good reporter and I actually specifically put the question to him. I said, "Am I not right in thinking that people understand the Fed’s statements since the COVID crisis as amounting to it, saying that it's unlikely to raise rates until 2022 or three?" He said that that's absolutely his perception of what the Fed has been signaling.

Selgin: So, it's out there, David, and I think that they have to be more careful if it isn't their intent to provide what seems like unconditional forward guidance even if it's only vague and doesn't actually give a definite date. I think they have and we know they said they're definitely not planning to taper and I want to know why that shouldn't be a data driven decision. And when I say they have no intention of tapering, I mean, for some time. They've given the impression it's going to be some time. Well, maybe it should. Maybe the data should be capable of telling them this month, next month, whatever. That it's time to start tapering. So, I see a tension.

I think that they have to be more careful if it isn't their intent to provide what seems like unconditional forward guidance even if it's only vague and doesn't actually give a definite date.

Share this

Beckworth: Okay. Well, one last comment and we'll move on and get to the Fed master accounts here in a minute. But I'm going to make a bold prediction here, George. You can timestamp it and use it to come back and tease me later but I think a decade from now, maybe two decades from now we will look back at the early 2020s and see it as the beginning of a price level targeting regime change for modern central banks or for advanced economy central banks. In other words, I think we will look back and see the early 2020s as being, for price level targeting, what the early 1990s were for inflation targeting. This could be the catalyst of change.

Beckworth: The Fed is the first central bank to undertake it, which is again, pretty ambitious, pretty impressive because before it was the Reserve Bank in New Zealand that did so and it was a smaller, little more nimble central bank. So the Fed is undertaking a huge task and the thing is, the Fed's doing it, the ECB has some officials who have said, "We need to follow suit." There's an article recently in the Financial Times that speaks to this, just last week, I believe. A policymaker, Olli Rehn, said, "Hey, the ECB should follow the Fed’s approach," and there's been a few other people at the ECB. So, the Fed does something like this. Let's just be optimistic here. Let's say it works, it becomes a version of a price level target. Other central banks begin to follow suit. This could be a very historic moment we are witnessing even if in real time there's a lot of noise surrounding other developments related to the pandemic and the recovery. Any thoughts?

Selgin: Oh, yes. Well, yeah. As usual I have to be the pessimist and not just because I don't think what you're saying may happen can happen. I'm a pessimist because I don't think that a concerted move to price level targeting is necessarily going to be an improvement. I know you and Scott have said that you consider it to be a kind of stepping stone between strictly forward looking inflation targeting and NGDP level targeting but I don't think that's correct. You guys know this. If there were no supply shocks that would be true but in a world of supply shocks price level targeting in the presence of those shocks is actually further removed from NGDP level targeting than plain old inflation targeting. So, it could be depending on how important supply shocks, including productivity shocks, especially are, it could turn out the NGDP isn't any more stable than in the past and could even be less stable.

Selgin: So, I'm not optimistic about this scenario because I think even if it does play out ... By the way, I'm not optimistic about living another 20 years. Or maybe I should say I'm not so pessimistic to think I'll live another 20 years. But anyway, I don't know that that would be a good improvement and I'd certainly like to hope that instead in 20 years, so the Fed might be taking NGDPLT targeting seriously because if it's not doing so by then, then surely we've wasted a lot of time and collective energy.

Beckworth: Oh, George, always the cheerful soul. Yeah, no that's a fair concern. I would just remind you though that what you mentioned earlier, the fact that the AIT is vague in certain dimensions. That could be a plus, right, in this situation. In fact, again, I'll go back to Ben Bernanke's description of a temporary price level target, which Rich Clarida said, "This is what we're trying to do." Now, that's Rich's version. But what Ben Bernanke said is, "Look, we want to do price level targeting when we've hit a real big demand shock that puts us at the zero lower bound but once we're away from that we want to revert to more regular type inflation target, which ..." And he says, "Sees through temporary inflation surges caused by supply shocks." So, in practice that could be very similar in spirit to nominal GDP targeting. But, you're right. Nothing's certain and it may not pan out the way that I hope it will and I certainly hope in 20 years, George, that our work is done on nominal GDP targeting.

Beckworth: Let’s switch gears here and move on to the big news that I want to discuss with you and that is this announcement by the Federal Reserve on May 5th. They put out a statement and it says, "The Federal Reserve Board on Wednesday invited public comment on proposed guidelines to evaluate request for accounts and payment services at Federal Reserve banks." And they go on and describe basically fintech companies would be the ones who would use it and access it and they say, "To help achieve a goal of applying a transparent and consistent process for all access requests, as well as considering the ramifications for their broader financial system, the board is proposing account access guidelines for the reserve banks to evaluate such requests." And you wrote a nice piece in The Hill and your piece was titled, *Keeping Fintech's Promise: A Modest Proposal.* So, let's start maybe with a real basic question. What is a fintech and why is it so important now that the Fed's considering this?

Fintech and Fed Master Accounts

Selgin: A fintech, the term is speaking of… The term is somewhat vague but I'm using it to refer to payment service providers that are not ordinary banks. Specifically they don't take deposits in order to finance loans with them. They don't engage in that traditional sort of bank intermediation, if your listeners don't mind me calling it that. For some people that's a dirty word. But instead, they offer other payment services, fast payment services. People are familiar with Apple Pay, with Zelle…

Beckworth: Venmo.

Selgin: Right. There's quite a few of these specialized payment service providers out there and they're all fintechs and in that aspect of financial services they are competing with banks. Now, things get a little complicated here and I know that my article confused some people but it's almost impossible to avoid the confusion in that there are now special depository institutions charters, or if you will, bank charters being offered both by some state governments and by the Office of the Comptroller of the Currency for fintechs. They're really designed for fintechs.

Selgin: So, you have fintechs that are banks in the legal sense of having charters, bank charters, and then you have fintechs that are not banks even in that limited sense, that legal sense and then you have banks, of course, banks that are not fintechs. So, there's really three categories out there. This Fed proposal is a request for comment that they're saying is only about how it should treat fintechs that have depository institution or bank "charters." It's only for them. It only concerns them. And that's important because it's less earth shattering than it might seem unless a payment service provider firm has a charter, a bank charter of some type, whether it's an ordinary bank charter or one of these special ones, it's not going to be eligible under any likely rules including the guidelines the Fed is proposing for a Fed master account. So, we should make that clear right away.

This Fed proposal is a request for comment that they're saying is only about how it should treat fintechs that have depository institution or bank "charters." It's only for them. It only concerns them. And that's important because...whether it's an ordinary bank charter or one of these special ones, it's not going to be eligible under any likely rules including the guidelines the Fed is proposing for a Fed master account.

Share this

Beckworth: But you would like that to be different, right? You'd like that to be the case where these other fintechs could also access the master accounts?

Selgin: Well, yes I would, David, and yet that's not what I was pleading for in that op-ed, which is even more modest. Because, here's the state of affairs right now. Although once a fintech has a charter, that makes it eligible automatically for a Fed master account. That doesn't mean the Fed has to give it an account. It's a discretionary decision at that point on the Fed’s part and in fact, the Fed has not granted any such accounts. It has had at least ... It has had several applications, I believe, which it has either rejected or sat on forever. It has one that I refer to in the article, in the op-ed, from Kraken Bank, that has been with it since last September but they haven't made a decision about it and there are no clear guidelines about well, how it should make such a decision. The Fed’s request for comment concerns guidelines it is proposing so that instead of being a black box that decision process will have some specific rules connected to it so that a fintech that applies can have some idea whether it's requesting something it's likely to get or not.

Beckworth: Well, let's make this a little more concrete by talking about that Kraken Bank. Kraken's an interesting name. That's a sea monster, I believe, from the Scandinavian folklore. But wasn't Kraken originally a cryptocurrency exchange or is that a different entity all together? Is this the-

Selgin: No, no that's part of Kraken Financial. Kraken Financial is a cryptocurrency exchange and it also offers up custodial services but as a cryptocurrency exchange, of course, that means it gets dollars and it has to buy other cryptocurrencies, et cetera, right?

Selgin: Now, it can do that more effectively if it has a Fed master account where it can then take advantage of the Fed’s wholesale payment rails to settle with other banks, these exchanges, purchases that have usually cryptocurrency of some kind as the thing being bought or sold and that's where Kraken Bank comes in. That it would be a bank to serve, that could handle the payments, exchanges, et cetera, that the business as a whole oversees.

Beckworth: Okay, so what you're doing then in that situation is you're cutting out the middle man. Kraken or any fintech would not need to go through a bank to have access to the Fed's payment system. Is that the idea?

Selgin: Exactly, yeah.

Beckworth: So, that would lower-

Selgin: It's a good point you make, David. Right now it is possible for any fintech to access the Fed's payment service without having a banking charter of any kind and without have to get the Fed's approval for a master account by hooking up with another bank. But since banks compete in this payments area it's a problem for fintechs to have to have a bank, to rely on bank partners and they believe they could do these things more cheaply without relying on them. So, there's a conflict of interest and there's an economy issue here. Both those things are at play.

Since banks compete in this payments area it's a problem for fintechs to have to have a bank, to rely on bank partners and they believe they could do these things more cheaply without relying on them. So, there's a conflict of interest and there's an economy issue here.

Share this

Beckworth: Yeah, so you would get cheaper financial services if this were allowed. But the other thing, just going back to fintechs, like Venmo and some of these others that you mentioned. The idea is where care about them because they're providing a good to the economy. Now, we mentioned Kraken. Kraken's associated with the cryptocurrency and some people may not be terribly excited about that but many of these other fintechs you can trace to get granting more access to people who've been unreached by traditional banks, right? Fintech can go where banks can't go and so this would be good for many people in the US who otherwise may not have access to the financial system. So, what we're discussing here is something that really shouldn't matter correct?

Selgin: Absolutely, yes. It's now, I think, pretty notorious that in large parts of the world including relatively less developed countries fintechs have provided great advances in payment services particularly for unbanked people. The classic case of this is mobile payments, which have taken huge strides in Africa and the Far East and a lot of them are done person to person or whatever and people can do them without actually having ordinary bank accounts and it's done a lot of good. So, for something like that you might want to have a telecommunications company offering the service where it had a bank subsidiary and when I say bank I mean, it's called a bank for regulatory purposes. It has a charter. It's called a bank charter or depository institutions charter and it could have a master account at the Fed.

Selgin: That's going to make it a lot easier for it to do what it wants to do efficiently. So, that would be an example. So, it's not just about cryptos and I should say, by the way, we should distinguish Kraken, for example, its crypto operations and the risks involved that are risks of holding crypto for the people who want to. That has nothing to do with its having a Fed master account. That account would be dollars for transacting purposes that clients of Kraken give to it in order to execute purchases and other transactions.

Beckworth: Yeah, let's talk about risk for a few minutes because we've been having some conversations with some of our friends and won't mention any names but the conversation got heated. There are very strong views on this topic. But what are the risk for these fintechs if they did get access to Fed master accounts? So, the way I would think about it. This is effectively like 100% reserve banking, right? So, all the fintech would do… you put funds into the fintech and those funds sit as reserves at the Federal Reserve. There's literally no way for it to lose your money. It's not doing maturity transformation, making loans, it's not exposing itself to risk the way a bank does and so, if that's the case, what are the risks? Why would we want them to, for example, to fund with some capital? Wouldn't the 100% reserves be enough? Why are people worried about them?

The Risks of Fed Master Accounts

Selgin: So, the best way for me to answer that, David, is to tell you the story of how I got into this discussion. Because I found myself sort of the man in the middle. I have friends at BPI and they, of course, represent the big banks and also, I know some people at Kraken or I got to know most of them as I investigated this subject. Kraken and BPI were very much at loggerheads here and Kraken had applied, as I said, for a master account with a Wyoming, a special purpose depository institution's charter it had obtained. So, it was eligible for a master account and as I noted, it applied but as yet hasn't got one. Well, BPI responded in a number of newsletters and publications saying, "This is a very dangerous thing. They shouldn't get a master account," and indeed I think I'm correct in saying that so far as BPI is concerned, these special charters are themselves a dangerous thing.

Selgin: What BPI pointed out and what they claimed, and I say claim… It's true but, and we'll see what the but is, that although Wyoming, for example, although its special purpose depository institutions charter has exactly the same, in fact, I think somewhat stiffer capital requirements than the clean old Wyoming bank charter, it doesn't require a special purpose depository institution like Kraken to have the FDIC insurance or any deposit insurance. Now, the FDIC has its own capital requirements, you see, and those are much higher than Wyoming's own requirements. So, BPI says, "Look, Kraken is risky and it should have to hold the same amount of capital as we banks do and of course, we have to hold plenty, in part, because of insurance and it's not fair for them to be able to compete with us with less of a regulatory burden if they're a risk." But this is where things get a little bit interesting.

Selgin: What the risk that Kraken could undertake under Wyoming's regulations is that it cannot make loans. So, that ordinary risk isn't there. That risk, I shouldn't say the ordinary. That major source of banking risk of borrowing deposits, which of course, are short term and lending long in loans. That's not there. However, Wyoming's law does allow SVDIs to invest in both federal government bonds and some corporate bonds of a certain safety. Well, BPI said, "That's still maturity transformation. There are still risks. It's interest rate risk but hey, that could be important and therefore, they should have to have as much capital as banks.

Selgin: Well, my response to that, and this is where I got involved, is two-fold. First of all, the risk is less so perhaps the capital requirements should be less too but more importantly, even though Kraken's charter allows it to invest dollar deposits in bonds of various kinds, Kraken itself said to me, when I asked them about this, they said, "Well, we could under the law but we're planning to keep all our dollar deposits in our Fed master account if we ever get one. That would be true 100% reverses." Well, according to BPI's own argument, if Kraken did do that then there would be no reason for them to hold this much capital. That's based on why BPI said they should hold more capital. That's if they engage in maturity transformation but with 100% reverses, that is reverses in the form of balances in a Fed master account, there's no need for that. There is no maturity transformation.

Selgin: So, that's when I came up with my proposal. I said, "Okay, well why don't we have a rule that says if you are 100% reserve backing fintech then first of all, you should get a master account. I think you should get one whether you have a banking license or not but certainly, if you have a bank charter you should and the Fed shouldn't hesitate because there's no reason to be concerned that you're having inadequate capital when the main reason you might need capital doesn't exist in the case of your business model."

Selgin: Now, there's another point to this and this is very important. I mentioned in the article but some people overlooked it. I'm not saying that anyone should take Kraken's or any other fintech’s word for it that they're going to hold all their money in a balance at the Fed, all their dollars, even if they don't have to. The Fed has the power today, and it mentions the power in its proposed guidelines, to hold any applicant for a master account to certain conditions and it could certainly hold a fintech to the condition that all of its dollar deposits have to be in its Fed balance. the Fed has the power to do that. So, it can enforce this requirement by proposing. It's not as if it has to just trust its fintech master account holders to abide by it voluntarily.

I'm not saying that anyone should take Kraken's or any other fintech’s word for it that they're going to hold all their money in a balance at the Fed...even if they don't have to. The Fed has the power today, and it mentions the power in its proposed guidelines, to hold any applicant for a master account to certain conditions and it could certainly hold a fintech to the condition that all of its dollar deposits have to be in its Fed balance. the Fed has the power to do that.

Share this

Beckworth: Yeah, that's really interesting and if a fintech, if Kraken did this, they stuck to 100% reserves, so there's minimal risk. In fact, what would be the risk, George? Let me step back. If Kraken gets access to a Fed master account, it keeps 100% of reserves. What are the risks that would face this firm. What could possibly go wrong that we should be worried about?

Selgin: Well, that depends on who we are, right? If you're a custodian at that bank. You've got a bunch of crypto stored up and you've given them your keys because that's what it means to have a custodial crypto account. You could still lose all your money. The firm could go broke but what can't happen is that somebody who has dollar deposits with Kraken bank can lose anything or that there can be any systemic problems because of Kraken's dollar deposits suddenly disappearing. That can't happen because as far as those contingencies are concerned Kraken is a 100% reserve institution and we only care about that.

Selgin: I mean, those of us concerned about financial stability, which should include the Fed, only care about any kind of systemic risk and the other thing to care about is whether dollar deposits are absolutely safe even though they're not insured and the answer to that is also yes, they are. So, that should be all. We don't care if Kraken bank fails as long as the other things are true and we don't care what risks it takes as long as it doesn't risk deposits, dollar deposits. So, this isn't saying that Kraken bank is a risk-free institution all together or certainly that Kraken, as a whole, is a risk-free enterprise. That's not true but that is not the Fed’s concern. It shouldn't be any financial…

Beckworth: Well, what this says to me is there's no way there'd be a run on fintech that had an account like this. There'd be no way to have a run on Kraken because there's nothing to run about. Your dollars were secured.

Selgin: You could run but you're going to get your money and…

Beckworth: Right, there's no reason to run because your money's guaranteed because the Fed is holding it and guess what? The Fed creates those dollars out of thin air, which leads me to another observation. What's the point of FDIC? Why should they get FDIC? FDIC is to protect deposits but your deposits are ... In fact, I would argue the Fed protection is better than FDIC protection.

Selgin: Perhaps, yeah. Well, it's better for the taxpayers for one thing. But, there is absolutely no reason, and therefore it's no wonder that a bank, a fintech, whatever you wish to call it, like Kraken would prefer a charter without deposit insurance being required because its business model is one where having such insurance doesn't matter if its business model is indeed a 100% Fed balance reserve model and like I said, they can be held to that. So, there's no need for that. There's no need for the high capital that banks hold because they engage in maturity transformation. So, you have a lot of politicians and regulators and bankers, of course, out there saying, "If it's a bank it should be regulated like a bank." But this is just playing with words. As I said, these banks are not doing… these fintechs that manage to get special bank charters aren't doing the same maturity transformation that banks typically do and therefore, they shouldn't have to be regulated in exactly the same way, though we can regulate them and should regulate them. There's an easy way to do so in most of these cases.

There is absolutely no reason, and therefore it's no wonder that a bank, a fintech, whatever you wish to call it, like Kraken would prefer a charter without deposit insurance being required because its business model is one where having such insurance doesn't matter if its business model is indeed a 100% Fed balance reserve model...these fintechs that manage to get special bank charters aren't doing the same maturity transformation that banks typically do and therefore, they shouldn't have to be regulated in exactly the same way, though we can regulate them and should regulate them.

Share this

Selgin: What I'm proposing with 100% reserves though is really meant to be a crack in the door. I don't want to scare the bankers or make it sound like this is some kind of Trojan horse. That's not what I mean. What I mean is this should be the minimum or the sufficient requirement for a fintech to get a master account but not a necessary requirement. We could also imagine other levels of risk where the fintech wouldn't have to have as much capital as a bank but would still have to have more capital than if it were keeping 100% reserves, and so on. So, I think that ultimately you want a regulatory framework that treats different kinds of payment service providers appropriately according to the actual risks they take and doesn't try to put them all in one unflexible category.

Beckworth: I think you make a great point. Like you said, you're just really identifying guidelines for maybe one extreme and there's degrees between that and a regular bank. But you could imagine a world where Venmo, you use Venmo where your dollars are being deposited at the Fed indirectly via Venmo. So, it makes a lot of sense to me and in fact, George, this sounds a lot like the Chicago Plan, right? 100% reserve banking. But with a qualifier that it's an option. You're not forcing the Chicago Plan on anybody.

Selgin: That's right, David. It's very different. I've had some people ... I got Morgan Ricks all happy about this because-

Beckworth: I was thinking of him. He would love this proposal.

Selgin: Yeah, and I'm glad Morgan is happy but I don't want to make all financial institutions 100% institutions. I don't want banks to have to be narrow banks as some people do. I don't want to get rid of fractional reserve banking. I've defended it all along despite how screwed up it's become in some ways. But I also recognize that if you are a fractional reserve banking institution you do need more capital one way or the other. Whether the market's going to make you hold it or regulation has to do so, as is the case these days because of guarantees. So, I'm not saying that fractional reserve banking is fine no matter how much capital is involved. I think we need more capital for firms that engage in risky fractional reserve banking and less for ones that don't.

Beckworth: No, yeah I agree with you. What I like about this is it presents options to the consumer. You can do traditional banking. I can go to Bank of America with the associated risk or I can go to Venmo who will deposit your funds directly in a Fed account for you and let people choose. Let the marketplace decide what's popular, what's not. But yeah, I was thinking of Morgan Ricks as we were talking and Lev Menand, are good friends who've advocated for Fed accounts and I think they would be very supportive of your position and in fact, it's kind of an indirect way to get at what they want. So, for those people who are underbanked, this would be a great way for them to get direct access to a Fed account.

Selgin: Well, that's very important, David, because I believe that if fintechs really had a proper, let's call it, despite the cliché, level playing field, a fair playing field here and what I'm proposing, I think, is a leveling step. They would be able to do more efficiently for the unbanked what others would propose to do by allowing people to have direct access to master accounts, individual access to the equivalent of master accounts at the Fed. So, I think this is a better solution and it's a proven solution elsewhere but that's perhaps another debate but I think that whether it ultimately proved to make Fed accounts for ordinary individuals redundant or not, this would be a desirable step for improving payments and access to payment services.

I'm not saying that fractional reserve banking is fine no matter how much capital is involved. I think we need more capital for firms that engage in risky fractional reserve banking and less for ones that don't.

Share this

Beckworth: Absolutely. I think what this also does is it satisfies, at least in part, Lev Menand and Morgan Ricks, their desire, but it also, I think, allows for innovation to continue flourish, to thrive. Right, if you've got competing fintechs, you've got banks competing. That's the big concern, at least one of my big concerns of granting access to Fed accounts to the public is that if the Fed becomes a retail banker for everybody, as well as an institutional banker for the big firms, it's hard to expect them to be innovating new financial products, new services. But if you've got fintechs out there competing and having access to the accounts, you can still see a world of innovation, new products, better services, lower costs, so forth, going forward.

Selgin: Yeah, that's right. And there's also a danger of the disintermediation risk and here politics comes in. Many of the fintechs I'm talking about, they would only ... Their dollar deposits would be as it were dollars on the move, right? Somebody's got to make payments. They use the fintechs to accomplish a quick payment. But the fintechs aren't their ordinary deposit holding institutions, those would be banks. But suppose a fintech is just trying to be a bank and accept deposits to place them in 100% reserves. If it earns the full interest rate on reserves that banks earn, there is some real risk of disintermediation, that is the banks will lose business to fintechs like that. This risk, by the way, is the one that has caused the Fed to balk when it came to giving a master account to TNB, The Narrow Bank, which is referring to a specific business. They were basically a bank for money market funds so the money market funds could earn interest on reserves for their clients when treasury rates were so low that the interest on reserve was higher.

Selgin: That kind of situation's very dangerous, of course. The Fed would have the right, at least under its proposed guidelines, to limit the interest rate payments on certain kinds of master account holders, non-banks, fintechs, and that could prevent this kind of disintermediation risk and the banks should be happy about that. They don't want to disappear. That's fine. But would the Fed be able to do the same thing when it came to giving it's own direct accounts to everybody, to ordinary individuals. Then I think politics would kick in big time and you'd have people putting pressure on the Fed to pay the full interest on reserve rate on those public accounts. It might not happen, but I'd be worried about it happening. And some proponents of Fed accounts for all, including Morgan, have said that they want the Fed to pay the full interest rate on reserves in those accounts.

Selgin: Well, okay but unless the account balances have a cap on them or limit that could be a huge problem because you could have circumstances where everybody just says, "I'm going to take all my money out of the bank and put it in my Fed account and goodbye banking system. That almost happened with postal banking accounts in the 1930s because they were paying an administratively set interest rate that turned out to be higher than what banks could pay while the depression was going on. Fortunately, those postal savings account balances were capped, otherwise, it would've been a huge additional source of disintermediation. So, there's some dangers here that people should be more aware of. By the way, I'd like to plug something. The Cato Journal just came out on our digital currency conference from last year and I have an article about this disintermediation problem with digital currency in this new issue.

Beckworth: Well, George, the time has flown by. I'm not able to cover everything I wanted to on the show with you but one thing I do want to cover before we end today is related to what we've been discussing and that's the payment system in the US and you were on the podcast with me some time ago and we discussed some of the developments and the big development was the announcement of the FedNow, the real time payment system the Fed was adopting and we talked about how there was already an existing one and how the Fed had created problems. And maybe if you could give us an update on that and the other thing I'd like an update on, and this is a critique you've made, I think, as probably the most poignant one, is that the Fed has existing, or what you call legacy payment systems, several of them. And the problem with them is that they don't run 24/7. In fact, they keep hours that sound like the old banker hour joke. They don't run that long even during the day.

Beckworth: I know that you and Aaron Klein over at Brookings have written about this and talked about this but if they would just fix the legacy payments it would go a long ways in making payments quicker and not take as long. So, maybe if you can give us an update on the FedNow real time payment development as well as if there's been any updates on legacy payment systems.

FedNow and Legacy Payment System Developments

Selgin: Well, David, I wish I had more of an update to provide. The only thing that has happened since the last program, if I recall correctly when we had it, is that the Fed is now saying that FedNow will definitely be finished in 2023 whereas before they'd been saying it might be 2024. And I haven't heard anything more since. That still is so late that by the time it's done the rival real time retail payment system run by the Clearinghouse called RTP, which was set up back in 2017, will really already be covering something like 80% of all deposits. And it could cover many more. Banks are joining RTP all the time because they can't wait. More and more people want faster and even instant payments.

Selgin: But to get to the legacy question. There's been no new development on that unfortunately. But this is consistent with what I think has been the Fed’s plan all along. It's putting all its eggs into the FedNow basket and it also, I think, is taking advantage of people's confusion about fast payments. You see, payments can be really slow like they are now in many cases where you can have some days go by before a payment settles and they can be super fast as they are on RTP, instant almost, or as they will be on FedNow but there's a lot of faster payment options in between. That is where you're not instant. Your payment isn't instant but it's not taking days and those are the improvements, those are the possibilities the Fed hasn't done a thing about and is giving people the impression they don't exist.

Selgin: And the main thing there, as you said, is that the multi day delays in settlement are on those payment networks that rely on the Fed’s wholesale settlement systems, Fedwire and the National Settlement Service, which are closed on weekends. Now they do have pretty good weekday hours but not until very recently, March 21st, where those hours themselves, those weekday hours extended to the point where you could get a one-day settlement using the Feds facilities even if you were making a payment in California, right, because of the time difference.

Selgin: Well, now they fixed that. Now they have three times a day when ACH payments can be sent and completed in the same day and that's a big improvement but you still have Saturdays, you still have Sundays and you still have holidays when those services are down. And the Fed has talked about doing something since, I think, as far back as 2012. It's been encouraged to get more extensive hours but it, as far as I know, hasn't done anything about ... It hasn't moved an inch toward trying to open those facilities on weekdays and holidays and that's a shame. It's not that it's without technical challenges. There are always challenges. It's not as simple as just being open. There are some problems with timing and the system has to be down at certain times, et cetera, but it's certainly possible for them to do it and it's a heck of a lot easier than what they are doing, which is FedNow and most people don't care about the difference between instant and within a day.

The Fed has talked about doing something since, I think, as far back as 2012. It's been encouraged to get more extensive hours but...It hasn't moved an inch toward trying to open those facilities on weekdays and holidays and that's a shame.

Share this

Selgin: For the kind of people Aaron is worried about who are waiting for their paychecks so they can pay their bills, that's not the delay difference that matters. It's the days. Finally. It's not the case that when FedNow is up and running or when RTP has got everybody potentially connected or both… It's not the case that these legacy systems are not going to continue to operate. Payments on them may be cheaper and for many kinds of payments, the FedNow and RTP systems, they don't work. They don't work, for example, for getting your direct deposit wage. That's still going to be done on ACH. There's all those people Aaron's worrying about that they may still be relying on the ACH system and probably many will. There'll be some alternatives for switching over but not complete and therefore, the Fed needs to fix this other thing. It should've made it the first priority and then only worried about FedNow if indeed it bothered with it at all given that there's another system that can do what it's going to do.

Beckworth: Well, it's remarkable they haven't done more, given all the attention you and Aaron have brought to the issue.

Selgin: Well, they're not the only ones who don't listen to me. I don't know about Aaron.

Beckworth: Well, they threw you a bone, it sounds like, by extending the hours during the weekday but the weekends are very important as Aaron ... He's been on the show before. We'll provide a link to that as well. But he notes it's pivotal between making payments for many of these people. So, it's surprising. Let me ask one last question and we'll end our show here. If Kraken, if Venmo gets access to a Fed master account, does that solve these payment problems for people?

Selgin: It could help but-

Beckworth: It could help, okay.

Selgin: It could help but not really because you still have the problem of ubiquity. Ultimately you want everyone connected to one network. Now, and the advantage of the legacy networks is everyone with a bank account is connected, right? And the advantage of RTP and FedNow will be the same advantage but faster payments, whereas with other fintechs, most of them will be providing better payment services for their own clients or for smaller networks that are not encompassing the entire set of people with bank accounts, let alone everybody.

Beckworth: Okay. With that our time is up. Our guest today has been George Selgin. George, thank you for coming on the show again.

Selgin: Anytime, David. Anytime and I want to get another mug, by the way.

Beckworth: Okay, I’ll talk to the bosses at Mercatus. Alright, take care.

Selgin: Bye bye.

Photo by Daniel Slim via Getty Images

Support Mercatus

Your support allows us to continue bridging the gap between academic ideas and real-world policy solutions.Donate