Banner for 2025 Financial Markets Conference

May 19, 2025

Policy Session 3: The Evolving Market for Bank Funding

The third policy session of our annual Financial Markets Conference explores questions around the evolving market for bank funding. The panel, moderated by The Brookings Institution senior fellow Nellie Liang, includes Comerica Bank executive vice president and treasurer Stasie Kostova, CFA, Davis Polk & Wardwell LLP partner and head of financial institutions practice Margaret Tahyar, and NYU Stern School of Business professor of economics Lawrence J. White.

Transcript

Raphael Bostic: Good morning, everyone. There was a little bit of learning from yesterday. Good morning, it's good to see you all. Welcome back, and welcome to all of you who are joining us via the live stream.

I hope you all enjoyed yesterday as much as I did. It was an excellent blend of learning for me, as well as a little bit of relaxation—I probably should have squeezed a little more of that in there—and I hope and trust that it was the same for you. Yesterday, I really enjoyed my conversation with Vice Chair Jefferson, and I particularly liked getting up to speed on the ins and outs of how nonbanks are showing up in financial markets, and all the different ways that's playing out. And I have to say, my brain is still wrestling with the issues associated with AI that arose in the context of last night's dinner conversation, and I actually have a feeling I'll be thinking about that conversation for quite some time.

We have another stimulating day on tap, and the schedule will feature presentations and discussions of timely issues that will help shape the future of monetary policy, global financial markets, and the macro economy. Now, everyone here has the agenda, so I'm not going to go into detail on it. I will just put a particular flag in the ground for the evening conversation that we will have as a capstone. It'll be my pleasure to share the spotlight with San Francisco Fed president Mary Daly, as well as one of the newest Reserve Bank presidents, Beth Hammack of the Cleveland Fed; and knowing them as I do, I'm sure that they will not disappoint in having a lively conversation. I'll just try to keep up with them as much as I can.

I hope everyone enjoys the day. Please engage, continue to participate, and ask questions. And now, I will welcome Nellie Liang, who will moderate today's opening panel. Nellie?

Nellie Liang: Thank you; thank you, President Bostic, and the Federal Reserve Bank of Atlanta, for having us here today. I very much appreciate the opportunity to speak at this session, which is: "The evolving market for bank funding." And this fits into the sessions, the whole conference program, incredibly well; it's timely. It will focus on banking models, non-banking models, and even new technology, probably not AI, but we will get into crypto, to some extent, with stablecoins.

So, this session is about, is low-cost stable funding still that for banks, or has it become more expensive and less stable? And there are a couple reasons why: it could be non-bank competition, it could be technology, as consumers demand different types of services.

And just as background: banks' deposits provide at least two services. One is liquidity for consumers, and another is a payment instrument for consumers. And this is one of the core reasons for bank franchise values. Unlike credit, which we talked about yesterday, where credit-to-Gross Domestic Product (GDP) for banks has been declining for the past few decades, deposits-to-GDP has not—it's been rising. This is really, in my view, a core feature of banking; but they are facing greater competitive pressures from nonbanks and from technology, and so that's what we're looking forward to discussing today.

And so, we have a great panel here today to discuss these trends, and I'm just going to start on my right (just very brief introductions; they're in your book): Stasie Kostova, who is the executive vice president and treasurer of Comerica. We have Meg Tahyar—did I get it right?—a partner and co-head of financial institutions at Davis Polk; and then Larry White, professor of economics at NYU. And you're probably familiar with all of these panelists today.

So, let me say, we're going to organize our conversation around three areas: first, what's happening to low-cost stable funding for banks? What have we learned from the bank failures in March 2023, and the stresses that were observed at other regional banks? And are there new vulnerabilities in the business model to be addressed?

Second, what have bank supervisors, regulators, and banks done in response? And do we need to be doing more? And then third, we'll talk about how banks are responding to increased consumer demand for digital payments and peer-to-peer payments, including stablecoins, and how they might affect the stability of bank funding. And then we'll turn to the audience Q&A.

So, I will start with our first area, where we're just kind of establishing the baseline (I'm going to turn to Stasie first), which is: Can you provide, from the banking sector, a perspective on banks' access to low-cost stable funding, and how important are these new developments from nonbanks or technology, Stasie?

Stasie Kostova: Well, thank you, Nellie! It's great to be here. So, competition for bank deposits is not new. We're used to it in our industry; it has always been intense. However, as Nellie alluded to in her question, there have been some powerful trends in the industry that are impacting the competitive landscape. For one, the move to digital has been underway for a while now in the industry, and it's really driven by customers' demands for convenience in accessing banking services.

So, we have seen banks rapidly digitizing. They're offering online platforms, mobile apps, a wide range of payment products—and that has the effect of making it very easy for customers and businesses to move money between different institutions. So, in essence, that reduces the friction that used to make deposits more stable in the past.

However, this also has a positive side. These new technologies are allowing banks to access new deposit opportunities. For example, online platforms allow banks to reach beyond their physical footprints. This move to digital has also invited competition from fintechs into the industry, and as you know, fintechs can compete with banks in certain areas such as payments, investments, et cetera. These are all activities that can be sources of stable low-cost deposits.

At the same time, we're seeing an increasing number of examples of banks partnering with fintechs in strategic Banking as a Service (BaaS) arrangements. This again, is offering opportunities for banks to reach new deposit customers, who would be normally out of reach for banks; some examples include gig workers, small vendors, et cetera.

So, fintechs can be a competitor for banks, but they can also be a very good partner. And fintechs and banks can collaborate sometimes in very powerful ways, to really take advantage of both sides' strengths: on the fintech side, fast innovation, cloud-based services, and on the bank side, direct access to payment rails, deposit insurance, compliance with bank regulations, et cetera.

So, these trends are really introducing new risks in the industry, as well as providing new opportunities for banks to grow their deposit franchises.

Liang: Larry, any comments on bank funding models, or per the perspective of your research—your very voluminous research—in the banking area?

Lawrence J. White: Sure. My sense is, as Stasie said, access to the low-cost, basically stable deposit environment is still there, but if things start to get rocky for a bank, the flightiness of deposits has increased substantially. That's what we saw with Silicon Valley Bank (SVB) two years ago, and in that sense, the overall system is more fragile. Even a modest-sized bank might be considered systemic now if its depositors start to run. And then the normal (normal?) contagion process of others at other banks noticing the run at the first bank, and they start to get nervous.

And that, of course, is heightened by the fact that the percentage of uninsured deposits in the banking system is substantially above what it was, say, 30 years ago, when I had to deal with the savings and loan debacle in the late 1980s. The amount of uninsured deposits in the savings and loan system: 2 or 3 percent. We didn't think about runs problems; we had lots of other problems, but not runs problems.

And so, as of 2023, the percentage of deposits that were uninsured was 40 percent. It's gone down a little bit since then; but still, that's substantial, and that makes the whole system much more fragile. The flightiness and—again, the technology that enhances that flightiness; under normal circumstances, things are going to be fine, but things can get worse a whole lot faster than we used to be accustomed to.

Liang: Thank you. Meg, do you want to comment on this?

Margaret Tahyar: So, Larry, I absolutely agree with you, that this technology has meant that things can go faster, quicker. We lived through 2023, and a lot of small businessmen, a lot of charities, a lot of religious organizations that quickly go over the 250—they lived through it too. So, they're now multi-banked in a way they weren't before, and it doesn't matter that they're not in Silicon Valley, that they're not techie.

I have this character that I use when I teach law—Joe, who has 100 auto body shops. He graduated from high school; he built this company from scratch. He's all over Southern Michigan, and I'm telling you that—

White: You didn't say he's got grease under his fingernails.

Tahyar: He's got grease under his fingernails, but I don't say he's not techie, because he's using high-tech machines in his shops. But he's not international. He's not networked. He's not social media. He's not any of those Silicon Valley things. Joe had one bank on March 1st, 2023. By the end of March, because there are no flies on Joe, he had more than one bank. And next time, Joe is just going to swipe right and leave, as you were describing, Larry.

So that, for me, the lesson from 2023 is, there will be a next time, and it's going to be faster. And there's all kinds of specialness about what happened to SVB and First Republican Signature—which, the Stephen Kelly paper that's recently out, is really good on—but nonetheless, next time, Joe is going to be participating.

And just one slight comment, Larry, in what you said: I think the 40 percent figure, if I remember correctly, is for mid-size and regional banks, and then when you go into the large mega banks, I think they're even more uninsured. But when you go to community banks, I think you're looking at 70–80 percent. The bottom line is the same as what you said. It's just that given the structure of our banking system, different banks are placed differently here.

Liang: Would you like to comment a little bit more on what technology—Meg pointed out social media, and social media has been hyped as a big factor, perhaps, in SVB. Meg mentioned the Stephen Kelly paper—there are others that have pointed to the role of social media, and maybe some doubts about how important it is in the sense that social media is often really concentrated at the retail sector, and private individual depositors—whereas most of the runs at these banks in March 2023 were by large corporate and institutional investors.

So, there's some doubt as to the role of social media, and therefore what supervisors might or might not want to do about it. But there are other technologies, and so this idea that you will be multi-banked: how much pressure is that putting on the banking system, to have—all of a sudden you have much more competition, because of this ability to switch?

Tahyar: I think that's partly why we're seeing the rise of brokered deposits, the rise of reciprocal deposits. An interesting thing is reciprocal deposits are mostly used by banks $100 billion or less; they're not used by bigger banks. Lots of people can have lots of views on brokered deposits, which I think are traditionally—and Larry, you would say, were "hot money" maybe, from back in the savings and loan days (although now they tend to be time). But I just look at the increase in both of those as sending us some kind of signal.

White: And of course, there are the new categories of entities that issue short-term runnable liabilities, and stablecoin-type entities that offer more competition, no question. They may be small now; the same could have been said about money market mutual funds in 1973 or 1974, and we know where that went.

Liang: Thinking about what are the sources of vulnerabilities for funding, and while banks have long managed this kind of competition—these are standard business models, and banks of all...there are some new features, and technology. Large uninsured deposits, perhaps large uninsured and concentrated deposits, at some regional banks in the regional bank sector—the other parts of the banking sector were not as affected—and the role of the partnerships with fintech and whatnot.

So, if we take that, given what has been changing, what actually have banks been doing to respond to these changes? What have supervisors and regulators done in response to March 2023? And I'll put parenthetically, if anything—because I have this bias that not a lot has been done, and maybe not a lot needs to be done. I think that's a very open question.

But I just want to turn to that, and again I'm going to start with Stasie. How are banks responding to these developments—especially, let's say, of March 2023? Because you spoke about the long-run evolution, but how are you dealing with customers with multiple banks?

Kostova: So, customers with multiple banks—that's not a new phenomenon in the industry. That was the case before the events in 2023, especially if a customer is large and sophisticated. We normally see them multi-banked, they usually have syndicated facilities on the lending side, they will have multiple banks on the deposit side. So that's not a new phenomenon, but we certainly have seen an uptick as a result of the events in 2023.

In terms of the events in 2023, I do want to talk about the importance of deposit insurance. It's a point that Larry made as well, and Meg: deposit insurance is very important in the competition for bank deposits, and for that reason, I think we need to have a substantive discussion about ways to modernize it. When we think in general, deposit insurance, generally speaking, has been effective when it comes to retail accounts. However, in its current form it has a big blind spot for commercial deposits, because of their size and complexity. And for that reason, the vast majority of commercial deposits are uninsured, across the entire industry.

So, what we saw in 2023, as regional banks and smaller banks were trying to level the playing field with the "too big to fail" banks, they were indirectly trying to increase insurance limits for their customers, particularly commercial clients, by turning to syndicated deposits, such as reciprocal deposits. In my opinion, a more direct recognition of the needs of these important deposits will be more effective and more sustainable in the long term.

We saw in 2023 the Federal Deposit Insurance Corporation (FDIC) publish a report, a detailed report, on deposit insurance, including some ways to reform—potential reform—and one of the suggestions was to expand deposit insurance in a targeted way to operating commercial accounts. I'm a supporter of that approach, and as you know, operational commercial accounts are those deposits that companies keep with their banks to support core business functions such as vendor payments, payroll, et cetera. So, these deposits, in essence, are non-speculative core deposits that serve a very important economic function.

So, this proposal, in essence, is to remove the systemic perceptions, while preserving accountability in the system. Now, I do realize that this will require congressional action, but I think we can at least start the conversation. I also want to caution that if we continue to ignore the needs of these important deposits, we will continue to see larger concentrations of commercial deposits, particularly in times of stress, into a handful of very large banks, and that can lead to systemic risk. So, I think that's a very important thing to discuss and try to resolve.

And just to wrap my thoughts: In order to have a level playing field in the competition for bank deposits, we need to have a conversation about ways to modernize deposit insurance, and to recognize the needs of all deposits, including commercial deposits—particularly given how important they are in supporting core business activities that help the entire economy.

Tahyar: Can I expand on that? Just the same theme. We're probably in raging agreement here. In 2023, Maxine Waters and J. D. Vance, then a senator, both put out bills that would do this. So, I think that's a pretty strong signal in our divided world, when those two people are suggesting a similar policy solution.

Now, some of the things that have been said about—what do we call them, operational deposits?—the deposits that some kind of entity uses on a daily basis, right? It's essentially the checking account that is used for daily payments, when it's not in its fancy clothing. I think it's not just commercial deposits. I think it should be 501(c)(3) charities. I think it should be religious entities. If you're a religious organization—a church, or a synagogue, or a temple—and you have 30 employees, you're above 250.

And those were the deposits that were fleeing really very much out of the regional and mid-sized banks during March 2023. And I think if we want to encourage something that doesn't create a barbell in our banking system, we're going to have to engage with deposit insurance reform. But of course, it's not even number 100 in the agenda of Congress right now, and so we need to really have that conversation.

Now, something that's often said is, it's too difficult to know what an operational deposit is, and I just want to strongly push back against that. It's not difficult at all. You'll get a statute that will give you something general. It will delegate to the regulators. But there's really a very easy, functional, record-keeping kind of way, which is just to say: let's just call it a "beanie" account, okay? If you meet the requirements, you're a beanie account, and your account would add—somebody has to pay for it, so there'll be fees—but let's just pick $5 or $10 million, that it is insured.

In the documents that open the account, it has to say this is a beanie account. And then in the technology that keeps the records in the bank, it has to say this is a beanie account. So, if the bank fails, the FDIC can come in and the computer will generate a list of those accounts that are covered. You're not deciding it at the moment of failure. You put the special "beanie" on it at the moment of opening, and if it's not put on, it's not covered.

So, I think people of goodwill could find solutions here, if there was political will.

White: Let me go even farther—and I think Jonathan Swift, the 18th century satirist in England, would recognize at least the title: I'm going to make a modest proposal of 100 percent deposit insurance. And let's see, is anybody saying the words "moral hazard" out there?

Tahyar: Is anybody saying, "Is there any other country that does that?"

White: Look, we should be a leader in many ways, and this is one of them. Moral hazard is not the issue. Depositors have never been good monitors of banks, never. And so, with 100 percent deposit insurance, runs are history—which would be great. And we talked about small banks; it gets small banks on a level playing field.

And bank regulation is known to be very opaque, very hidden. Why? Because the regulators are afraid that bad news will lead to runs. That goes away. We can make bank regulation a lot more transparent.

Tahyar: You mean regulation or supervision?

White: All of the above; sorry, supervision. I keep on forgetting that there's that distinction. So, much more transparent, and just...there are so many good things that go along with this.

Now, that will make even more clear the responsibility to have better supervision, better examination. That's a good thing as well; and you mentioned earlier, what about the examination and supervision? Clearly, there was a shortfall in 2022 and early 2023. We've got to make it clearer how important that is, and also give them better information, which means mark to market accounting. That's got to be the prominent way of information that supervisors see—not as an afterthought, but the upfront information that they see, and also just as a regular basis for revelation of information.

Once a quarter? Now, I understand supervisors get much more common information, but as a general matter, once a quarter? Come on, that's a 70-year-old technology. We ought to be thinking about once a week, maybe even once a day. That's the way investment banking used to conduct its information, and maybe even real time. What a thought.

We've got to think more proactively, in terms of better regulation, better supervision, a better examination, better information, and more frequent.

Liang: Okay; so now we've gone to the extreme levels of reforms, which is a great point to make. Just one follow up, before I go back to Stasie. Mark to market: is this mark to market of securities, or also loans? Where's the private value of a bank involved?

White: Right now, we've got the mark to market of securities.

Liang: Well, not exactly. That was one of the problems.

White: It's not part of the main body of information. It's in a footnote, but any anybody who's familiar with Silicon Valley Bank knows that as of the middle of November 2022, when the 10-Q for the third quarter became public, all you had to do was look at—I forget, it's page 18—and there were the numbers that showed that Silicon Valley Bank was, on a mark to market basis, insolvent. We need that information to be even more prominent and provided more frequently.

Liang: Okay, I'll come back to that. Stasie, you wanted to mention something?

Kostova: Just very quickly, to respond, this is part of a substantive discussion. There are different ways to approach deposit insurance reform. The FDIC report also offers different ideas. The definition—going back to the operational deposits definition—I believe that problem can be solved. Banks already have a definition of operational deposits in the liquidity coverage ratio (LCR) framework. It's not uniform across the industry. I think that can be solved. Regulators can provide a uniform definition and apply it across the industry.

And we have plenty of precedents where regulatory definitions have to be followed and applied by banks, so this will not be a unique problem to solve. But I welcome this discussion; I think this is part of bringing different ideas that will address the problem.

Tahyar: I just want to respond to Larry. There's another thing, instead of being the only country in the world that goes to explicit uninsured deposit; there is a place in the middle, which is to bring back the program and the powers that the FDIC had, before Dodd-Frank took them away, which was—and I may be getting the acronym wrong—it's a Transaction Account Grantee (TAG) program, right? So, we have fractional reserve banking. It is what it is. It's the original sin of banking. We haven't been able to move away from that.

So, if we're keeping that system, at least for now, what the TAG program did—and this is why we did not have deposit runs in the Great Financial Crisis. I'm going to repeat that: we did not have deposit runs in the Great Financial Crisis. The reason we didn't is because of the TAG program. It permits the FDIC, with some guard rails and charging a fee, to make all what they call "non-interest-bearing transaction accounts" unlimited insured, for a limited period of time (we would call those "operational deposits" now). That power was taken away from the FDIC after Dodd-Frank, as a reward for having done a good job during the financial crisis, but I think that's because Congress was just shocked at how much power the Fed had, and the FDIC was kind of innocent citizen and collateral damage.

It was secretly put back in during COVID-19. A lot of people don't know that, but the Coronavirus Aid, Relief, and Economic Security (CARES) Act put it back in again, for a limited period of time. It was just never used. So, this is a turning-on-and-off-in-periods-of-stress-with-a-fee valve. It's worked in the past. It could work in the future.

Liang: So, there have been a bunch of other alternatives to full deposit insurance, some recognizing that a large expansion of deposit insurance could be costly, so, I just want to just push on those just a little bit. Larry, you mentioned you could see that Silicon Valley was insolvent, so, what is the role for capital regulations? So, that's one question.

A second one is: There have been proposals to improve access to the discount window, to be operationally ready to pledge collateral and get funds. And in fact, after SVB, there have been public reports of how the supervisors required banks to be prepared to access funding from the Fed if they needed it, to reduce the likelihood of further deposit runs.

So, there are two others; one is operational—improving discount window capacity. A second thing could be, how do you transition from home loan bank advances to discount window? I know as a banker, Stasie has a view on some of these. And then, capital regulation. Those are other options on the table, so would you all like to comment on any of those?

White: Let me add, there's yet another, which is: reduce the runnability of those uninsured deposits, add some friction in some way; put a penalty on a quick withdrawal, or insist—essentially, reduce the liquidity of the uninsured deposits. There have been a number of proposals; Jeffrey Gordon, a corporate law professor at Columbia University, has, I think, a very interesting proposal there. My colleague, Dick Berner, has also advanced an idea.

The basic problem here is that the quick withdrawal of an uninsured deposit may send a message to fellow depositors, to depositors at the bank across the street. It's basically a negative externality phenomenon, and we've got to deal with that negative externality in some way. If we're not going to have a 100 percent deposit insurance, which would be my first preference, then think about that negative externality penalty.

As an economist, I think, put a tax on whatever it is that generates the negative externality—carbon tax, for global warming-type issues. Put some kind of friction that makes it harder—not impossible, but harder—for that uninsured deposit to quickly leave, and you tamp down that contagion problem.

Tahyar: I know we're not there yet, but when we get to stablecoins, I think we should discuss how stablecoins might impact that idea.

White: Right.

Kostova: All these suggestions, they have advantages and disadvantages, like everything else in life. So, these discussions are helpful. We also need to think about preserving banks' competitiveness. We're not just competing with each other, we're competing with other participants—mutual funds, fintechs. So, that's also an important consideration, in considering some of these approaches.

Tahyar: I'm wondering, Larry, I'm thinking about your idea of the—and I'm leaving stablecoins and money market funds and all that aside for now—would you impose that on all uninsured deposits at all banks, or would you let depositors opt into, "I'm willing to have a gate like a money market fund in return for which, one gets a higher interest rate?"

White: I'd have to chew on that a little bit more, but it sounds like a good....

Tahyar: I don't know if it's a good idea, it just occurred to me. But you said "tax" and I was thinking more in the private sector capitalist mode of, "Well, wait a minute. If you want someone to take more risk, you need to pay them more."

White: Okay. Yes, that sounds right. I'd have to think a little bit more about it, but—

Tahyar: Fair enough. I think just moving back to the bank runs—we've got 800 years of bank runs. We've got 800 years of proof that depositors are not good moral hazard monitors. So, I kind of lose patience with the moral hazard argument for depositors. Now, from the 90s, we had a lot of thinking about subordinated debt as providing a market signal. There was a lot of back and forth on market discipline.

But the big lesson of Silicon Valley Bank (and its famous footnote showing that it's insolvent) is, why didn't more people notice? We've got a lot of asset managers in this room, right? Some asset managers noticed and left. Where were the equity analysts? Show me the equity analyst report in the back half of 2022 on Silicon Valley, where there was a bank equity analyst, someone who specializes in banks, who pointed out that Silicon Valley Bank was balance sheet insolvent. That's a failure of market discipline, and it's almost a failure of the intermediaries in the system.

So, that gets to your point, Larry, on more transparency, because had there been more transparency—and I'm not quite sure where the line is here, but on the supervisory side— had there been many, many more people seeing more information about banks and what's happening, maybe someone would have noticed earlier. Because the cost of the diff, had it gone down in November 2022, would have been much less.

White: Yes, for sure. And again, I really believe that supervisors ought to be seeing the mark to market, fair value, market value—whatever terminology you want to put in place there—see that as the basic information. If there then is a footnote that says, "Oh, by the way, generally accepted accounting principles (GAAP) historical-based accounting shows that Silicon Valley Bank is solvent." Okay, fine, but the upfront, in-your-face information, I'm an economist, I believe in markets and market value accounting. That's just got to be the direct information.

And by the way, Nellie, for all, not just for securities, for everything. Yes, that's going to take some modeling. There are some smart economists who've already started doing that type of modeling. Again, it's never going to be perfect, but it can be done.

Tahyar: So, two responses about two very different points you're making. The other point about Silicon Valley Bank and the famous footnote is, did the supervisors read it? Do the supervisors read the cues in the case? What training does the supervisory staff have on public company disclosure, and the reading of those footnotes? Because there's nothing in the bank's annual report (BAR) report that talks about when the supervisory staff started to understand that there was this horrible, horrible interest rate risk problem.

Moving back to loans. A very, very different point you're making, Larry—and on this one, I would need more time to think about it, but I think that if it were possible to mark loans to market, that it would have been done before (but I'm willing to be talked out of that position). We had the session yesterday on private credit, and all those medium-sized enterprises and the middle market credits, and that market is—there's limited information on it, but how do you mark to market a loan to that company? A Fortune 50. I can see it. A small- and medium-sized enterprise, or a medium-sized company—really hard.

I can even see it in consumer, because you can do portfolios; but in terms of what the banking sector provides to this country that is fundamentally important for our economy (and I'm not an economist, so...). It's all of that credit that goes to Joe when he was starting out. And if marking to market impacts that, I think we need to have that as part of the discussion.

Liang: So, we have a couple of questions from the audience that I'm just going to insert as we go, because I think there's quite a bit of interest in thinking about deposit insurance here. So, we touched on reciprocal deposits and the reliance of some banks on those, et cetera. With a 100 percent deposit insurance, I assume that business goes away, right? It's just an arbitrage on this. But let's say we don't do 100 percent deposit insurance. Let's just say, Congress would never get there. How do you think of reciprocal deposits? Are they entirely destabilizing? Are they just all cost to banks, or is there some positive value to them?

Kostova: My experience is that reciprocal deposits are very stabilizing to a deposit base, and it's an indirect way for banks to increase deposit insurance limits. Keep in mind, these are the bank's customers that are being reciprocated through these FDIC sweep networks. So, these are core customers to the bank. The bank has a relationship with them. It's a tool for banks to really provide a higher level of insurance. And because, as I spoke about the insufficient level of insurance, particularly for commercial accounts that are large and complex, that's a tool for banks to really stabilize their commercial deposit base. It has been very effective.

Tahyar: There's a great paper out. It's quite recent, from Ned Prescott of the Cleveland Fed, on reciprocal deposits. He has a co-author—I'm sorry, I don't remember the name. Somebody can shout it out if you know it—but it's a fabulous paper, and it really walks through reciprocal deposits, who's using them, how they're being used, and I recommend it toad it to your precious, limited reading time.

Liang: So, another proposal came as a question. One of the facilities, the Bank Term Funding Program, was put in place following SVB, which took Treasury securities at par to the Fed, and lent at a one-year rate (I believe it was one year). It's just for information. It did require Treasury to put capital in this facility, so it is not a lending facility that the Fed can do on its own. It did require some capital, but that is one way to approach mark to market on securities.

So, the question is, what do you think of this, Larry?

White: Look, everybody in this room knows that when, under normal circumstances, someone lends against securities as collateral, there's a haircut, okay? What the Fed was doing was the exact opposite: ignoring the mark to market on the Treasuries. My colleague, Kim Schoenholtz, described that as not a haircut but a toupee. I wish I had been that clever. It just goes so strongly against the idea of paying attention to the actual value of the securities.

Liang: Right. So, it was a crisis management tool, not a pre-emptive tool, to prevent runs, exactly.

White: Something needed to be done, that's for sure, but there had to be—look, the Federal Reserve has a lot of very creative people. They could have found a better way to tamp down the danger of runs than putting a toupee on those marked securities.

Liang: I agree; the Fed has very many creative people, and the smart work people who address...but this was an issue of contagion, to address across the system. And as I said, it isn't a Fed program. It was a Fed and Treasury program. It does require capital—taxpayer support. In the end, it didn't have to pay out, but you didn't know going in. But it was an effective part of the response to prevent contagion in the system.

But again, on a preemptive basis, you would create a different program. You would create a different type of facility if you wanted to be...you could look ahead, but in the moment, this is what you need to do.

Tahyar: There were other things, Nellie, and you just alluded to this, that were revealed in 2023. Neither Silicon Valley nor Signature Bank were able to borrow from the Fed. The reasons were largely operational. There's a lot—and we've heard a lot in this session, and we've heard from the Fed and also from the Federal Home Loan Banks, about a lot of work that's being done to operationalize things so that it's different next time.

I have a feeling, however, that we're still kind of midstream in developing that system. I don't think we're quite right at swipe right, swipe left—whichever swipe it is—to get collateral from the Federal Home Loan Bank into every single Federal Reserve Bank in the country.

And I think—I don't think, I fear—that there's still grit in that system. So, I absolutely encourage the folks who are down there in the coal mines working on those operational bits to know and understand that what they're doing is super important. And I'll remind folks that the Federal Home Loan Banks are massively over-collateralized, and they have valued the collateral; and part of the grit in the system is the Federal Reserve Banks need time to value the collateral. But if there was a little bit more trust—and you've got a Federal Home Loan Bank that has already valued the collateral, and is 300 percent over-collateralized, and the Fed would be 150 percent over-collateralized—can't there be a little bit of "hands across the waters" here? I'm not quite sure why we have to have different systems of valuation.

I just think that operational work—we hear a lot about it at a very high level. I think it would be comforting to the market to know and understand that there are a lot of people working in those coal mines, and then to get reports on what's happening with actually making it operational.

Liang: Just to comment on that, and just to give a sense of the operational complexities: there are 11 Federal Home Loan Banks and 12 Federal Reserve Banks. So, the original mapping starts to get difficult, but we get it, we get it.

So, a different question is, just following up on the mark to market, which is kind of like the discount window—you can kind of go down a rabbit hole. But I do just want to raise this issue: Would you also mark to market liabilities? So, do you mark to market the whole balance sheet, or do you also think about the liabilities when you have this proposal? I think that's always been a very tough question, if you have any thoughts on that.

White: It is a tough question. Look, there's a valuation, say, to having a stable deposit base. That ought to get valued in some way—except, as we've learned, that value can evaporate very quickly, so you need to be very careful about how that marking works.

But in principle, yes, I'd be prepared to certainly think about marking the liability side; but you've got to be very careful on that.

Liang: There are a number of questions about regulators, which I think I'll come back to because I think the next session will also raise this. I'm going to go to the next topic, which is about payment innovations, which includes stablecoins—which are a different source of risk to low-cost stable funds.

We know consumers value deposits as a payment instrument; they're increasingly using digital payments. Actually, a survey by the Federal Reserve Bank of Atlanta—I think it's called the Survey of Consumer Payment Choice—has documented, especially since COVID-19, that the use of cash is way down, the use of credit and debit cards are up, but the use of digital payments and app usage is way up, and peer-to-peer payments are growing.

So, COVID-19, plus technology, has changed the way consumers are approaching payments. Stasie, you mentioned collaboration with fintechs or competition with fintechs, and how you're working with them and competing with them on faster payments. Maybe you can offer a little perspective on what you've been seeing, and how the banking system is thinking about this right now.

Kostova: Yes; from a broad industry perspective, I talked about the trends of digitalization and banks rapidly modernizing their platforms—we see this in the area of payments as well. We can see a wide range of payment products and services being offered by banks now, and I'll just give some examples: as you know, Zelle is a service that banks use to allow customers to move money between accounts, into different institutions. More banks are starting to offer digital wallets now. We're seeing that banks are increasingly adopting instant payments, which includes the clearinghouse, real-time payments, as well as the Federal Reserves' FedNow.

Banks are also modernizing their platforms to support APIs and cloud-based services. I mentioned that banks are getting into strategic alliances with fintechs, in terms of "banking as a service" arrangements. So, there's a wide array of services, payment services, and products that banks can now offer to their customers.

Liang: Can you just explain a little bit of the "banking as a service" arrangement? This is the BaaS acronym?

Kostova: Yes; it can be a very powerful partnership between fintechs and banks, where it really plays to the strengths of both sides. On the side of the fintech, it's fast innovation, cloud-based services, they can sometimes reach customers that banks don't have access to (I mentioned examples with gig workers, small vendors, et cetera).

On the side of the bank, it's really giving direct access to payment rails, offering deposit insurance. Banks really have mature bank regulatory compliance and capabilities, so it's really the best of both worlds—where customers can access advanced payment options, while having the security that the bank can bring.

Liang: I see.

White: Let me just add, though: essentially, what that is doing is, instead of vertical integration, it's vertical disintegration—which is fine, which works if there is the comparative advantage at those various stages. But the regulatory structure has to take that into account. The "know your customer" requirement has to be extended to that third party, the anti-money laundering—all of that.

When the bank was vertically integrated, then you relied on the bank to adhere to all of those. The bank outsources those functions; you've got to make sure those requirements are imposed appropriately.

As far as stablecoins are concerned, if that is a service that customers want, fine. But I think we need to avoid a situation that got created with money market mutual funds, where there they are issuing short-term runnable liabilities—but having no capital, and none of the other kinds of requirements that the banks, as issuers of short-term liabilities, have to have.

So, we don't want to replicate the money market; if they're providing a useful service with the stablecoin, fine—great. But they've got to have capital requirements, they've got to have other liquidity requirements, other kinds of bank-like regulation applied to them.

Liang: So, I guess let me just say a couple of words about stablecoins, as we're getting into this part of the discussion, and then bring it back to bank funding and the stability of bank funding.

So, you're in luck. Congress, last night—the Senate—voted to remove a procedural hurdle to passing stablecoin legislation, and I believe more than a handful, maybe two dozen Democrats, voted for cloture.

So, what are stablecoins? I think, to just back up for a minute, I think the way to think about it is, it's digital cash—just cash in digital form that is offered by a private firm, not by the central bank, or necessarily a commercial bank. It's not our system of money; it's kind of outside our system of money. It's going to allow transactions to be cleared and settled instantaneously, 24/7—basically—on a blockchain, and so that makes it different from cash, paper cash as we do it, or from like just a digital payment with a credit card.

And the technology is different. It's going to be more like Zelle, but Zelle is not a blockchain-based payment. So, the current regulations—and we can come back to a few questions from the audience—the current set of regulations for stablecoins is, state-level money transmitter laws. So, big stablecoin issuers have a money transmitter license from all the different states, maybe multiple licenses from some states. If you actually are going to operate a payment system that tends to cross state borders and crosses national borders, you can see the kinds of complications this could raise, especially if a stablecoin issuer were to fail and you tried to resolve how to make sure the payments happen.

Currently, most stablecoins are probably used for crypto trading; I think that's been documented. There's some limited use in the world for sort of real economy cases in the United States, but it is being used by other countries that often look to US dollar-based stablecoins as a store of value. And so, some of these countries with more volatile domestic currencies, or that lack a strong, coherent financial regulatory framework, may look at US dollar-based stablecoins as a store of value, and they use this. That's actually been where a lot of the growth has been, and then it's also been used for cross-border payments and remittances.

It's grown a lot, but it's not huge; when I was at Treasury and we wrote the President's Working Group's report on stablecoins in 2021, stablecoins then were $20 billion—but even at $20 billion, I don't know where we had the confidence to say this: we thought it could become a financial stability risk if you didn't address the runs (this is kind of like the early money market funds). Now about $250 billion, it's been growing despite the ups and downs of the crypto business. So, where it'll be in five years is hard to know; a lot will depend on how banks respond to more digitization, more tokenization—they could create bank deposit tokens.

But it is growing, and so there are some in Congress now who believe some kind of regulatory framework is helpful for financial stability, for illicit finance, and for supporting the global role of the dollar, and national security reasons. But those all raise this issue: if Congress, again, passed a procedural hurdle to consider a stablecoin legislation further, the House has a bill. They're not exactly the same, but they go through reconciliation and the White House is very much pushing for stablecoin legislation this year. It does feel like the potential for it to grow, at least in the near term, is much higher now than it was a month ago.

So, how do we think about that, in terms of, what else do we know about stablecoins? I'll open it up to this panel. How do we think about that in terms of, does it change funding for banks, or change access to low-cost stable funding for banks?

Tahyar: Let me start. First of all, thank goodness Congress is doing something, because we have a new financial instrument/payment mechanism that is out there in the wild in nature, and there are all kinds of policy issues that need to be worked through—at a congressional level, and then at a regulatory level. So, we often criticize our congress for being a do-nothing congress; now they're doing something—and they should do something.

But we should also say, thinking back in the past: they're not going to get it perfect the first time around, and there's going to have to be some regulatory iterations, and then there will likely have to be some changes. So, I don't have a crystal ball for what stablecoins are going to do. There are a lot of what ifs out there in the world, and a lot depends on what happens.

So, just a couple of figures that I think are interesting. Currently, stablecoins are at about 1 percent of the $18 trillion of US bank deposits. At the moment they're minuscule. There's a Citi report out there that's kind of interesting, and it says that their base case is that they will be $1.6 trillion by 2030, and their bull case is $3.7 trillion. Now, that's an analyst prediction.

They, however, in transaction volume are already about (this trailing last 12 months) they're about the same size as Mastercard and Visa transaction volumes. So, they're a thing; they're becoming a thing. What's going to happen next? I've got no crystal ball. I want to be humble, as we kind of think through what's happening, so I don't want to make any predictions, but the possibilities are, the banks may respond with tokenized deposits. The banks may respond with higher yielding deposits. The banks themselves can go into stablecoins, and I think a number of them are doing that. They can be custodian for the reserves.

Obviously, one of the key innovations in this bill is it doesn't prevent Walmart or Apple or Google, or any company, from going into the stablecoin business. Also, one of the unsolved things in the bills is, what do we mean by "novel applications of the Bank Security Act (BSA), anti-money laundering (AML) regulations," and how is that going to work? So, there's a lot to work out, but the choices aren't looking forward to a future of what ifs and might ifs, and doing nothing; because if we do nothing, we're just going to continue to have this kind of beast in nature growing, without putting it into a regulated environment.

I think one of the more interesting things in my life, you've got a whole sector who's asking to come into the regulatory perimeter, and that's very interesting.

White: I'm glad you mentioned Walmart, because—as you know—there's a history of Walmart wanting to get into banking. They tried to do it through a Utah industrial bank, couldn't get deposit insurance from the FDIC, and backed away. But they've been eager to...they've done various kinds of financial services, check cashing, certainly allowing local banks into their lobby, et cetera. But I find it interesting that they could actually get into this kind of depository.

Tahyar: That's from the late 90s, right? So, maybe some people don't even remember, but there's interesting academic research that the reason that Walmart wanted in at that time was to bring down its interchange and payments cost. As folks in this room know, we've got the highest payments cost, much higher than Europe and in other parts of the world—India and China, because they didn't have a system, so they built something modern. So, I think that Walmart's goals back then, as I understand from the academic research, were to bring down payments.

And then, as a matter of fact, in reality, they kind of did get into banking; they own a bank in Mexico. And if I ask my students, when I teach as an adjunct, I say, "Is Walmart into banking?" And of course, none of them were born in the 1990s, so they say yes—because when they walk into a Walmart, there are all these banking services.

White: But not if it's a local bank having a branch in the store.

Tahyar: It's a partnership.

White: It's not the same.

Tahyar: It's banking as a service, before we had that name.

Liang: I just want to follow up on Walmart and big tech: so, when—again, going back to the President's Working Group (PWG) report, and the whole reason the US and central banks around the world began to worry about stablecoins, was when Facebook (at the time; Meta, now)...Facebook at the time introduced Libra, which was a multi-currency-based basket, a new digital money based on a currency. Then they switched it to Diem, which would have been US-dollar-based.

But the concern there is, as we know, money and payments have these network externalities and can scale up pretty quickly. And so, if you combine that with an online platform, which has access to data about all kinds of things about your spending patterns and your sleeping pattern, whatever; and that it could network and create an internal ecosystem of payments and money, and there were concerns about losing control of the money supply.

So, that was one of the three—there were three recommendations in that report. One was, make sure the value is stable. Two, make sure it can actually operate as a payment instrument (operational). And three, separate banking and commerce (the way we usually talk about it is separation of banking and commerce). That provision, that last one, has been extremely contentious on the hill right now. I think they have a modified separation right now, but it is definitely a contentious issue; and where it has settled is not clear, but if you do lobby and want to have a say in this, this is a key piece of information.

There was a question about, should they be regulated like money market funds? Just like government money market funds: should we regulate them? I guess I'll just start, and then I want to let the group start.

While their backing is like a money market fund, although the current legislation allows a little bit more, they are not a tokenized money market fund. So, the stablecoin is like a payment instrument, and money market funds aren't really used for payments. And they don't pay yield—they don't pay a yield, and so they're slightly different.

And the current regulation would have them be sort of regulated like a narrow bank. They do one activity, they issue a stablecoin, they redeem a stablecoin, they manage their reserves—that's kind of it. But there would be some capital and liquidity regulations, so they are distinct from money market funds, and so the regulatory framework that's being set up is a bit different—I want to be clear on that. Not to—again, this is just proposed legislation at this point.

White: Again, I'd like to see a capital requirement in there, which money market mutual funds have managed to evade. And nevertheless, they get bailed out, periodically, when they break the buck and can't redeem, and all of a sudden, we've got a contagion problem. You've got to have capital and other bank-like—

Tahyar: Just to be clear, Larry, but I think there isn't capital in the current bill. Can you remind me?

Liang: There is.

Tahyar: There is capital. How much is it?

Liang: It's up to the state and federal regulators to determine how much capital that would be needed to support the ongoing activities of being a stablecoin operator—of being an issuer. So, it is not set by the legislation.

Tahyar: It's something to—if you've got one-to-one reserves, which is the same as you have in money market mutual funds, you probably—now we have prime funds, mostly...sorry, no—government funds. I'm in the opposite way; government funds, mostly.

Liang: Many of us in this room have worked on that for a long time.

Tahyar: So, I just think it's going to be very interesting to see at what level that capital is set, and how it is set—just because that creates a bit of a stressor where you can encourage or discourage stablecoins if you're setting the capital too high. And I understand if you're setting it too low, then you've got stability issues.

But on the other hand, for stablecoins—and again, I don't know about the future and how it's going to play out, but does it matter in stablecoins if folks have to wait a little while for their money—in the way that it truly would matter if it were operational, and it has mattered in the past if it's money market funds? I don't know.

Liang: I think the presumption is, it would if it's a payment instrument.

Tahyar: If it's a payment instrument.

Liang: Yes, and that's what the legislation is for. So, we did have a question here: How would they be dealt with in bankruptcy? And coming back to our favorite topic: Should deposit insurance be extended to stablecoins?

Tahyar: So, those are two very different questions. As I understand the bills, and the reason why the stablecoin issuers are in separate subs—even if it's a bank, it's in a separate sub—is to create a bankruptcy-remote subsidiary. And as I understand it, there are some changes in the bankruptcy code as part of this that will make it clear that those reserves belong to the customers, and are not part of the larger organization.

So, I think if I trust those of my bankruptcy partners, they think that that's been mostly solved but not yet seen in the real world. I would come down on no deposit insurance for stablecoins. I'm sure Larry's going to say if you do that, you have to do it for money market funds. But it doesn't—if we're going to let different pathways bloom and try them out, we can't just extend deposit insurance to everything.

One of the biggest things since the financial crisis is that the switch is—who got bailed out in the financial crisis? It was bondholders, and uninsured depositors. I think we've taken a social view about uninsured depositors in most bank failures. Now bondholders know, 15 years after the financial crisis, they're at risk; and that's what we saw in Credit Suisse, and there was disclosure but people forgot. There was disclosure in Silicon Valley Bank, but people didn't notice it.

We've got to find some way in the world to make the disclosure around stablecoins, the disclosure around money market funds, something that people don't forget. And maybe the lessons of Credit Suisse will help people not to forget.

White: Any time you've got significant quantities of runnable liabilities, you've got to be concerned. And that calls for some capital requirements, other kinds of requirements, and examination and supervision to go along with that. My instincts are, do I really want to extend deposit insurance outside of the banking system? But if I'm concerned about runs, if I'm concerned about contagion, I have to think about that possibility.

Tahyar: I think it's too early. If you think about where stablecoins are, I just think it's too early to be thinking about things like that. We should create a regulatory framework, bring them into the perimeter, get both those who are stablecoin issuers, those who are the reserves, the supervisors, the regulators—let's get a few years of experience, and then decide. I just think it's too early.

Kostova: Yes; and from a practitioner's standpoint, that's an important point is to make sure that we're not in a situation of regulatory arbitrage, where different players providing similar services are exposed to different requirements. I think it can distort competition and affect the banking industry.

White: Absolutely. Arguably, that's the story of money market mutual funds. It was a regulatory arbitrage.

Tahyar: Larry, an interesting thing about money market funds is, they were just created accidentally, by some Department of Labor interpretation.

Liang: Yes, it's a security, not a deposit—something like that. Okay, I'm going to let one more final question for everyone, just to give you an opportunity; you can either answer this question or just mention something you wanted to mention and didn't get a chance to. It's a general question about following the bank runs and more fragility of bank funding: Has enough been done, either on liquidity or capital regulations? And what would be the one thing that hasn't been done that you think needs to be done—other than deposit insurance, because I think at least this group thinks that that the extension of deposit insurance is a positive thing. So, you can either answer that question or just comment on anything else you would like to wrap up with.

Maybe I'll start with Meg.

Tahyar: Alright. I think the one thing that hasn't been done that ought to be done, is the reform of supervision, more transparency around supervision. And I have a lot of respect for very fine supervisors who do their jobs, but we need more transparency, we need more accountability, we need more understanding on their training and curriculum. I think that the banking agencies ought to be putting a lot more out there in terms of what they're doing, so that scholars and investors can really understand how supervision works.

Liang: Should we consolidate the banking agencies? Let me just follow up on Meg for a second. You don't have to answer this: Should we consolidate the regulatory banking agencies?

Tahyar: In a perfect world, we wouldn't have the system we have. We live in this imperfect vale of tears. The academic in me thinks that we should consolidate them, but I know that that is so politically off-piste that the best we can do now is try and have what I'll call functional consolidation around, say, singing from the same songbook, to quote the Treasury secretary. Also just trying to have fewer overlaps and more consistency across the regulations. Why the heck are the confidential supervisory information (CSI) regulations different among the three regulators? So maybe if we could get a little bit more efficiency, we could maybe kind of get there.

White: I'm not a big believer in that kind of consolidation. I think there is value in differing perspectives, differing places somebody with a good idea can go. If he or she is turned down here, maybe that good idea can get accepted over there. So, I think one has to be careful about consolidation.

The answer to your question: No, not enough has been done. I would have liked to have seen the higher capital requirements and greater liquidity requirements that the Michael Barr efforts turned out not to yield. What's the one thing? Market value accounting.

Liang: Assets and liabilities, or just assets?

White: Market value accounting.

Kostova: So, again, from a practitioner's perspective, thinking back to 2023, I think away from the very idiosyncratic issues related to Silicon Valley and Signature, banks demonstrated that they had effective liquidity risk management tools. They were able to manage through the industry turmoil. The one thing, again, from reality versus what's reflected in regulations, is the current liquidity regulations don't recognize fully how banks manage in a real crisis.

One example is, banks have reliably and consistently, in industry stress, accessed the Federal Home Loan Bank (FHLB) system—and that's based on loan collateral that is prepositioned at the FHLB, as well as securities collateral. In each industry stress, banks have reliably and consistently accessed the broker deposit market as a source of diversification of their funding. So, these are practices that are not recognized in the liquidity regulations, and the result is that we have to substitute that liquidity in our liquidity buffers, by holding higher levels of cash and security.

So, every time you increase those requirements, it's less available for lending.

Liang: Yes; very helpful. So, thank you very much. I think we covered what we planned to, and maybe more; so, I appreciate the comments from our panelists, and please join me in thanking them.