Steven Kelly on Crises, Stability, and the Fed’s Role in Financial Markets

Read the full episode transcript:

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

David Beckworth: Steven, welcome to the show.

Steven Kelly: Thanks, David. Great to be here.

Beckworth: It's great to have you on. Now, I have listened to you multiple times on another podcast, your friend, our friend Kaleb Nygaard has The Reserve Podcast. And it seems like you're on every episode with him, or most episodes with him. So you're kind of like the co-host over there. So I encourage listeners to check that out as well if you haven't already, The Reserve Podcast. And it's been great to chat with you online. And I've been following your Substack, what is the name of your Substack so listeners can know?

Kelly: So it's called Without Warning. The main reason for that is just that, I had to have a title. And the other reason is that it's a little wonkish and I don't really have a cadence. So they come without warning, and I also just kind of skip the intro part. I can only write in March 2020, there was COVID, volatility, so many times. So I kind of skip over that when I write stuff over there.

Beckworth: Okay. But is the Substack free at this point? And so I'd encourage listeners to get it if you are interested in financial crisis discussions. And one of the articles, actually a couple of the articles from that Substack is motivating our discussion today. And that relates to the role the Federal Reserve plays, and whether it should be even more expansive than what we've seen. But before we get into that Steven, you are a part of the Yale Program on Financial Stability. So walk us through that. What is that program? What's your core vision? What are you trying to accomplish with it?

Kelly: We're kind of a lean mean bucket shop within the Yale School of Management. And we were founded in the aftermath of the GFC by Tim Geithner, who is our chair. And really the main motivation is, the story of 2008 is crisis fighters started feeling their way around in the dark. Ben Bernanke talks about how, when he got to the Fed he asked for a crisis playbook and it was four pages of open the discount window. So basically we are trying to fill this wartime consigliere type role. And we've heard from a lot of our counterparts in other institutions that it would take a lot of political capital in some of those institutions to do the work that we do, which basically presumes a financial crisis. So there's a lot of great work out there on MacroPru and forecasting crises. And we start with the assumption that all failed and what is the practical way to design interventions?

Kelly: So the bulk of our work is in what we call the New Bagehot Project. And we're really Andrew Metrick's brainchild. He runs our day to day. So we have the New Bagehot, which is backward looking, in that we look at every intervention documented throughout history and see what worked and why. And really we try to get into the weeds of implementation and contracts and this kind of thing so that crisis fighters in the future can pull it off the shelf, put white out over the name and write something else on top of it. And so there's a sort forward looking thing too, where we spread the good news of how to fight crises. And we have an online tool so that's out there as well as. You want to compare and contrast different interventions and terms that's out there. We have a master's program for people involved in financial stability, central bankers, Treasury officials, traders, who's ever got a mandate for financial stability. Very intense program, but you get a master's in systemic risks. So, that's kind of a few of our main pieces.

Beckworth: Yeah, I'm looking at your webpage for the New Bagehot Project and there's options to click case studies, surveys, more broad topics, lending buying. So you can drill down into whatever specific niche that you have, if you're fighting a crisis. So if you're sitting as the chairman of the Federal Reserve, Jay Powell one day wants to think about an issue. He can click on a couple of links and there you go. So this is a tool for central bankers, Treasury officials, to deal with financial stability concerns. And your point is it's, given the crisis is already there. What do you do? Right? That's the context as opposed to the preemptive kind of preventing it, so. Interesting tool, a lot of interesting work. And I also should mention you guys have a journal, right? The Journal of Financial Crises. So you take articles related to financial crises. Tell us about your work. What are you doing there?

Kelly: So, I'm predominantly focused on 13.3 and the US interventions. We cover, in theory, the whole world. Some countries have better documentation than others. But, so I'm primarily focused on the US and the Federal Reserve.

Beckworth: Okay. So you are focused on section 13.3 of The Federal Reserve Act for listeners who don't know. I think most of our listeners do know what 13.3 is, but for those who, don't the part that gives the Fed the authority to act in emergencies. And so that's a big part of what I want to get to with you today as well. One last question. So the Yale Program on Financial Stability, you deal a lot with, at least in the Bagehot Project, how to respond to a crisis once it has occurred. Anything else you're doing, any other preemptive or forward looking work as well, or it's just mostly after the fact work?

Kelly: We have conferences on, what are the lessons from these past interventions that we can carry forward into the future. This year I know we're going to be looking a lot at this idea of SME lending, and now climate change is kind of an issue. So we're thinking about all these things kind of on a go forward basis and really how can you arrest them in real time?

Beckworth: Well, the reason I ask that is kind of self-serving here. So I'm going to bring up what happened in 2020. And we'll get back to the main discussion in a minute. But one of the things we saw was this massive intervention in macro policy, combination of the Fed and the Treasury – fiscal and monetary policy – dispensing lots of money into the economy. And by 2021, in my view, it was too much, but in 2020 it was needed. And that income, that stable nominal income growth prevented a far wider financial disaster from emerging, right. People could make their mortgage payments, their auto payments. So, having a system, a macro system in place that provides stable aggregate nominal income growth, to me, seems like a great preventative way to avoid many financial crises. Of course, this naturally leads me to nominal GDP level targeting, which is another way of doing this. Now you might want to argue it needs to be complimented with fiscal policy and a severe outcome, that's fine. But in general, keeping nominal GDP or aggregate nominal income on a stable growth path seems like a great way to promote financial stability. So, if you guys get a chance to have a conference on that, I want to talk about that. I'd love to do that sometime.

Kelly: If NGDP targeting can put me out of work, I think that's a win for the world.

Beckworth: Definitely. Okay. Well, let's move on into what the Fed can do in crisis. And you've written several interesting pieces. Some articles I'll give a few of the names, but I want to step back and talk about the bigger point you're making in them. But one of the articles is titled “The Fed As Derivatives Dealer of Last Resort?”, and another one's titled “Could the Fed Rescue Commodities Markets?”. So just the name itself implies the Fed could go into places that it hasn't gone a lot in the past. I mean, it has gone into Maiden Lane. There was some derivative exposures you note in this. But in general, I get the sense from your writing that you think the Fed could do more of what it did in 2020 and in 2008, is that a fair interpretation? And if so, what's the philosophy behind it?

The Fed’s Role in Niche Financial Markets

Kelly: I think perhaps the way to put it is that the role is evolving as markets evolve. So, some of this stuff can read as expansionary for the Fed. And we talk about, oh, there's no market the Fed won't bail out at this point. And that's an ever present risk that we absolutely need to fend off any moral hazard and any incentive problems there. But I'd argue that a big piece of this too, is really the aperture of crises has shifted over time. You can think back 120 years, we had a crisis every harvest. I mean, we didn't even have that figured out yet.

An episode of financial instability, a crisis is maybe not the time for the Fed to pass value judgements on capital. If it's commodities dealers, if it's hedge funds, if they have capital that are willing to move the risk, it's incumbent on the Fed and its legal mandate to think about the liquidity needs in those sectors.

Kelly: So the aperture of crises has moved. Take 2008, for instance. It's not that the Fed never did anything like this before 2008 or that these rescues weren't happening, it's that they were happening inside banks. So you look at what happened over the course of 2007, bank leverage was effectively doubling as they're bailing out all their off balance sheet entities. They're trying to support the housing market. And if the crisis ends at the end of 2007, they're probably fine. And if banks go into 2008 with the balance sheets they had in 2006, they're probably fine. So we just used to do this inside banks and with monetary policy. And basically after 2008, we saw the downside of that. And we said, all right banks, that's it, you can't do this kind of stuff anymore. You got to include off balance sheet measures in your capital ratios, et cetera, et cetera. But that's means that the risks are going to be somewhere else. And someone else has to step in. I mean, banks can create money on demand when they do these intervention. Shadow banks, investment banks can issue money like assets. But outside of that, we're empty handed and you may have to go to the Fed.

Kelly: So, at the risk of sounding like I'm in Silicon Valley, it's a little bit of pushing the efficient frontier of what a crisis is, because banks are fine. I mean, what is the discount window now, aside from maybe a press release. I mean, we had 2020 and just this act of God in 2020 basically. And the discount window was nothing more than good PR for the banks. So, that's something that we kind of have to consider. I mean, we're trying to preserve the capital that's in the banking system. And so, an episode of financial instability, a crisis is maybe not the time for the Fed to pass value judgements on capital. If it's commodities dealers, if it's hedge funds, if they have capital that are willing to move the risk, it's incumbent on the Fed and its legal mandate to think about the liquidity needs in those sectors.

Beckworth: So in your writings you often make note of the growing role that market finance plays in the economy. So our economy is becoming more and more driven or funded by market finance versus banks. Is that kind of the big point here?

Kelly: Yeah. And it's like, where is the marginal risk going? So, we've decided it's a huge public good to keep the banks resilient in a crisis. And we've basically seen them being resilient. And so, even the quote unquote crises we talk about today, aren't financial crises in the same sense, where it's like the only way to stop this is to inject capital into the banking system. We're talking about hedge funds getting liquidated. We're talking about commodities dealers and commodities prices. We're talking about all these more fringe markets and that's an improvement. I mean, it's still a crisis. It's still a bad economic outcome. It might be unusual negligent circumstances, but it's not a financial crisis in the 2008 sense.

Beckworth: Okay. Well, I have some sympathies with your view, but I do want to push back and present the other side's view. And I want to draw upon a guest that we've had on this show several times, Lev Menand. He's a friend of the program. He was on previously 2020, I guess, right after the crisis started. He had a paper titled, “Unappropriated Dollars: The Fed's Ad Hoc Lending Facilities and the Rules that Govern Them.” And then he had a book just recently titled, The Fed Unbound: Central Banking in a Time of Crisis. And the way he looks at this is, he sees it almost as a policy driven emergent result of the Fed and in Congress allowing shadow banking to grow unchecked. So, the only reason he would argue that the Fed needs to step into these other markets is because they are issuing money like liabilities. So that means they can be run on, there can be crises.

Beckworth: And so for him, I guess that the central paradigm is, financial crises largely occur in settings where financial firms are issuing money like liabilities. And that's what shadow banking is, and we've allowed it to grow. And every time we step in, so 2008, 2020, all we're doing is reinforcing that tendency. So it'll be more shadow banking on the margin, adding more vulnerabilities. And thus, he looks at the growing Fed as kind of a result of that growth of shadow banking. And it's a vicious cycle, where you see it in a more benign perspective. The way we finance the economy has evolved, it's changed. And the Fed's just kind of keeping up with the new times. How do you respond to someone like Lev Menand?

Kelly: Yeah, I mean, Lev has great points. And I don't know that it's the Fed's role. The Fed can only allow controlled burns, effectively. It has a mandate for economic stability. And so, the idea that the Fed's not going to use a crisis to decide who can issue money like assets, to me that seems irresponsible. So if we want to reform shadow banking, I'm totally sympathetic. I don't know that a crisis is the time to do that. Because, I mean, these assets are serving macro needs, right? I mean, let's put deposits to infinity if we don't like repo and commercial paper and things like that. So, it's a challenge and that's always going to be the seed of a financial crisis, is going to be in something that resembled money. But a crisis is not the time to make that judgment.

Beckworth: Yeah, I shared that very pragmatic view, in the middle of the battle you don't want to start getting righteous. That, well, who caused this? Why was it there in the first place? You want to respond and stop the bleeding and then respond to it. And I think he would probably agree with that. He would just say, we need to do more when times are calm. Let me share a bit of my perspective on this. So, I look at what happened in 2020 and the Fed set up all these facilities. And I think particularly the global impact it had. Set up the dollar swap lines, the new repo facility for foreign official financial organizations, FEMA. And what that did, at least in my view, is it reassured investors across the world that they can hold dollar denominated assets and not worry about them. So on the margin, they're going to hold more dollar denominated assets. Now that may be viewed as a risk, but I also see that as an increased demand for dollar denominated liabilities, including US treasuries, US government debt. So some of that demand will go to privately issued dollar liabilities, for sure. But there'll still be some that goes towards the US government. Which means in short more senior ridge flowing into the US government. We can finance our deficits, easier, avoid any kind of situation where fiscal dominance emerges.

Beckworth: So in some ways the Fed’s stepping in actually alleviated on the margins some of the fiscal pressures are our country's facing. So I look at that as well, that's great. I also look at the global dollar system as providing the public good to the world. I wonder how much of the vast improvement of humanity coming out of poverty could have been done without a global dollar system facilitating trade and growth. And maybe there's another path that could have gone down, but there's a lot of positives as well. If you look at the global dollar system and keeping it intact. And that's where I come at this perspective, I know there's cost involved as well. And the final point I'll say, and I'll leave it at this. Maybe if you want to respond to it. It's not clear to me how we would fix a global shadow banking system. We could talk about doing it domestically, Lev has some suggestions. But it just seems to me, one of the key reasons you have shadow banking is because there's a whack-a-mole, you shut down one form of money creation over here and it pops up somewhere else. And so, I don't know if you guys have thought about that either. How would you reign in global shadow banking if you wanted to go down that route?

Kelly: Yeah, that's exactly the question. And that's the peace time question that we need to consider. And because the other thing is, sometimes this gets characterized as like, oh, it's regulatory arbitrage. It's just some fringe finance company issuing a money like asset. But as I alluded to with the question of deposit insurance, which is basically, the 250,000 thing is basically a myth that we all agree on. So, there's a question of, is this the best system, and it's probably not. And you've had Morgan Ricks on, and he's got some good ideas on this front. And just, I don't know if you can do it globally. It helps to have the global reserve currency. We've seen this to some degree with sanctions and things like that. Being able to control the dollar everywhere is a big deal, but I don't know.

Beckworth: Okay, well, let's move to the law. The Federal Reserve Act itself. And the part of that Federal Reserve Act that allows the Fed to respond in crisis, that allows us to do these very things we've been discussing. And that is the 13.3 section. So maybe walk us through that. And how has it changed? I know there's the big change after 2008, there was some interesting developments in 2020. And tell us where we are today?

Section 13.3 of the Federal Reserve Act

Kelly: Yeah, so this is the classic emergency authority of the Fed allows it to intervene in unusual exigent circumstances with the super majority vote of the board to basically lend to anybody, or discount for anybody against sufficient collateral. And there's a couple pieces to that, but basical ly the Fed can't expect ex ante to take losses, overall that is. So now the post-crisis changes, the big ones were, you can no longer do this for an individual. So the Bear Stearns Rescue, the AIG Rescue, that kind of stuff is off the table via 13.3. It has to be what they call broad-based facility. And so, the Fed has defined this as available to five borrowers. So, maybe in theory they could open a Maiden Lane to allow five people to use it.

Beckworth: So five's the number. I didn't know that. Interesting.

Kelly: So, the five number is not statutory, it's in regulation A. But they say that's their guidance, is we'd like it to be available to at least five borrowers. And I'm being a little tongue and cheek about the Maiden Lane thing, because the Fed does take the spirit of the statute very seriously. There's a lot of cynicism out there that, oh, the Fed will bend the law the way it needs to when a crisis comes. And that's not my read at all. It can seem that way from the headlines, we talk about them buying junk bonds and things like that. But they take this law very seriously. And the limits of it are very real. And they're not afraid to reject the many entreaties they get to use this authority. So anyways, post crisis, it has to be broad based. You can't lend to somebody in insolvency or to save them from insolvency. That's codified, the Lehman Brothers choices. That they have to be secured to protect the taxpayer. Again, this really didn't change much. And then the other addition, and Lev has talked about this, is that the assistance has to be for the purpose of providing liquidity to the financial system, whatever that means, we don't know. But that's another big change.

Beckworth: So Steven, they now have to apply a principle of broad based eligibility. Do they also need the approval of the Secretary of Treasury when they do it?

The last collateral constraint on the Fed's 13.3 facility [...] Basically what happened, this is just a couple lines in this bigger law in 1991 about federal deposit insurance. And basically what it is, Congress sees the volatility of 1987, Black Monday, and says, look, these securities dealers are a big piece of our markets now. It's silly to have this old codification that they are engaged in speculation and that's what they do. That's the service that they provide to us. That this is real market liquidity, it's real market financing, and we may need to lend to them in a crisis. So that was key to being able to lend to Bear Sterns and others in 2008.

Kelly: That's right. Yeah. So again, that was something that was practiced basically before, and it has now been codified. And we saw maybe a little bit of tension in 2020. Certainly towards the end of 2020, when some of the facilities got stopped that the Fed wanted to continue, and Mnuchin basically spearheaded ending them. But in general, there's a lot of cooperation. We'll get into, I assume, when the Treasury has some money involved, obviously the cooperation goes up anyways. But yes, this was codified, the Treasury now has to approve 13.3 in all instances.

Beckworth: So you note in your article that a big change to 13.3 occurred in 1991. Where they changed the language so that permissible collateral could include investment securities. So why was that such a big change?

Kelly: So, this is what allows them basically to do these market based interventions. So we think about the corporate bond facilities. We think about lending to Bear Stearns, lending to AIG. So this idea of preserving market liquidity – they specifically talk about market liquidity in the a company and banking committee report. What they say is, this clarifies that access to liquidity and special circumstances can be made available directly to a securities' dealer to help preserve market liquidity and avoid market disruption. So this was the last collateral constraint on the Fed's 13.3 facility. And basically what happened, this is just a couple lines in this bigger law in 1991 about federal deposit insurance. And basically what it is, Congress sees the volatility of 1987, Black Monday, and says, look, these securities dealers are a big piece of our markets now. It's silly to have this old codification that they are engaged in speculation and that's what they do. That's the service that they provide to us. That this is real market liquidity, it's real market financing, and we may need to lend to them in a crisis. So that was key to being able to lend to Bear Sterns and others in 2008.

Beckworth: That is interesting. So the 1987 stock market crashed, a real sharp decline. A lasting legacy of that is this change in 1991 to the law, which then today 2008 and 2020, allows the Fed to intervene in these market settings with the liquidity facilities. Huh. So we can trace a little line going all the way back to 1987. Interestingly Lev Menand mentions 1991 also is a key turning point, like in repo markets. The Fed was already intervening all the way back to 1950, he notes. But it's not until 1991, there's this big increase due to the change in law. And again, we can trace this back to 1987. So it would've been interesting to see a counterfactual history. No 1987 stock market crash, what would have happened? So very fascinating.

Beckworth: Okay, so you have that change in 1991. Let's go to the more recent past 2020. And I'm invoking Lev on in this discussion. That means I appreciate his work. He had in that first paper I mentioned, a distinction between liquidity facilities and credit facilities. And I actually like this distinction, liquidity facilities, you really are stepping in trying to keep liquidity, trying to keep the financial system running. Whereas a credit facility, the way he defines it, such as the Main Street Lending Facility, you are doing more credit allocation. And the danger I see with that is, you're starting to pick winners and losers. And I know that's going to be even true with liquidity facilities. You can make, oh, there's winners and losers there too. But when you get into Main Street facilities, it becomes more apparent, more pronounced, and it could raise political legitimacy questions about the Fed. And so, I'm wondering what your thoughts are. Is there a way to draw a line between the Fed getting over involved in credit versus liquidity support?

Kelly: Yeah. And this is where I think 13.3 Is actually a pretty elegant authority, because this differentiates the Fed from say the ECB or other central banks that are buying massive corporate bond portfolios as part of QE. So now they're engaged in all these discussions of, what is market neutral and should we green our portfolio and things like that. And the beauty of 13.3 is that it's somewhat insulated from that, in that, this is really about crisis time intervention. This is not about deciding on the future of the economy and what the future economy's going to look like. The fact that it's bound by unusual and exigent circumstances makes this effectively, makes them all basically liquidity facilities. I like the distinction, I get the distinction. But the fact that the board needs to find unusual and exigent circumstances, it has to find evidence that the recipients are lacking adequate credit. To me that suggests more of, okay, this is an emergency facility, and then it's going to be over. And by 2021, the Fed had sold all its corporate bonds. So, that's the distinction between like, oh, we're really extending credit and bailing out and deciding on the future shape of the economy, versus we're getting us through this crisis. And then our laws tell us to stop.

Beckworth: I totally buy that even for the corporate bond facility, because that you could tell a liquidity constrained story there. But when you get to Main Street, is there no, I mean, do you see no concern there? That maybe we're pushing the boundaries of what truly is a liquidity need versus, that they're just in a bad position? I mean, how long did that facility go on? I can't remember now. It went on well past the initial crisis period, didn't it?

Kelly: So, Main Street had to end at the end of the year. It was one of the CARES Act facilities that Congress wrote out. It was accelerating when it closed. But yeah, that one's tricky. And I don't know, because I don't know if the Fed even likes it that much. They didn't really like, the way they talk about it they didn't seem to it that much. It doesn't really serve the backstop role that market based type stuff does. It didn't do that much lending. And when you have no announcement effect, you kind of have to do the lending. And to me, it reads as a very direct response to the critiques of 2008. Which was, you're doing everything on three blocks in New York City. What about me? And I understand the Fed thinking about that. Section 13.3 very explicitly says, you may lend to any participant. So it is incumbent upon the Fed to think about, okay, let's think of every conceivable participant and who we might have to lend to and how we're going to design that. And when the crisis does call for it, and a pandemic, there's basically no moral hazard to go around. You've got a lot of people suffering. So I get the impulse, I think it's going to be harder for them to do in the future. And I think they're going to be less in a hurry to do so in the future, both from the performance we saw and just, it took a very high bar, I think, to even bring something like that out.

Beckworth: And I think there's a place here for another federal agency that could take the role of doing credit during a crisis. So I've had guests on the show in the past, in fact, Lev, and I talked about this. Having some kind of investment authority, kind of like the reconstruction finance corporation during the great depression that would play this role. And that way, there's a little more political legitimacy because Congress approves it, the people's voices are implicitly behind it. So, I think that's a safe way to go forward. And let me also just be very clear, even though I see the compelling argument for these facilities. I also understand the concerns that many have, that it increases the footprint of the Federal Reserve. Because once they're done, now you mentioned the corporate bond facility, it completely unwound and all the positions were resolved. But some people look at, for example, the Fed's large balance sheet is as byproduct of it intervening. And maybe that's a different issue altogether. But I understand that the concern that, is the Fed becoming in a bigger and bigger footprint in the financial system and the economy. But I think you're making the point you can delineate those two issues, or are they conflated?

Kelly: Yeah, I think it's two things. One is, what's a true emergency and do the legal authorities really bind the Fed to not be too interventionist. And then the other thing is, we need to think hard about the shape of our financial system, because it's not that these risks didn't, weren't always backstop it's that we had the banks in between the Fed and the risks and usually they could take it on. I gave the example of 2007, they're ending 2007 at 40-x leverage. That's the kind of thing. We said, we don't want you to do that anymore. We don't want you to take this risk because you're basically betting on a recovery that might not come. And if we don't let you do that, it's the same economic facts. But now with the Fed stepping in.

Beckworth: Well, let's take this discussion and apply it to some recent developments that you've written about, and the key one is the commodity market. So as we know, there's been wild swings and commodity prices. You note in your piece, even that oil prices went negative in 2020, pretty remarkable. Big swing in oil prices in 2008. And then today we're beginning to see some sharp drop in prices. So you have mentioned that the Fed could play some emergency role if things got bad enough, some kind of emergency role. So walk us through that. How and why would the Fed step in in the case of commodity market pressures?

Kelly: Yeah, so really the genesis of this piece was just, there was talk, particularly in Europe, about a potential government rescue. And I don't think we ever reached the point that it was necessary for the Fed to intervene. But I kind of wanted to get ahead of this question of like, oh, this is just the Fed bailing out BP. Is this legal? Does this fit within their authority? So I kind of just walked through the arguments. Unusual and exigent circumstances seems reasonable. We have Russian invasion, we have everything going on with commodities markets. Can the Fed be secured to its satisfaction? Lots of collateral. Lending against these agricultural type goods is basically classic central banking. We saw evidence of lacking adequate credit and all these things.

Kelly: And so I kind of just wanted to get ahead of it and say, again, this isn't a climate thing. This isn't the Fed lending to the oil industry. This is really about fixing the financial market problem. So we have that Dodd-Frank ad of, it must be for the purposes of providing liquidity to the financial system. This is really so the Fed is seeing a possible financial instability episode. It sees entities with capital willing to bear the risk. I.e., the commodities dealers, and what's strapping them is liquidity risk. So, it was just starting to think through those problems. And really, come to the assessment that the Fed could legally do this. We didn't quite get there. I don't think they should have, and things are calming down now. But there was an alternate reality in which some real intervention was needed.

Beckworth: Let's segue into your article titled “The Fed as Derivatives Dealer of Last Resort.” Because this is tied to the commodity question. And you note, the Fed has dipped its toe in the pool of derivatives back in 2008 with Maiden Lane. But maybe walk us through that episode as well as how it would work going forward.

The Fed’s Activity in Derivates Markets

Kelly: So again, it was just thinking about, how is the Fed going to take collateral in the commodities markets? And one of the things that's very difficult is making sure it's secured to their satisfaction, especially when things are so volatile. You mentioned oil going negative in April 2020. I talk about July to December 2008, which a barrel of oil goes from $145 bucks to $31. So, very hard for the Fed to be sufficiently secured in an environment like that. And a possible solution is that the Fed could take a derivatives position consistent with the so-called long position in commodities that they would be taking. Because a lot of these commodities dealers are, they have hedges in place and that's what was driving the liquidity demand. So yeah, the Fed has taken derivatives in the past. I guess first I should say, they can definitely take derivatives as collateral if there's collateral involved. If there's a loan underlying it, the Fed will take my TV as collateral if there's a loan underneath it. But there's a question of how much collaterals available in derivatives. If you write a futures contract at the money today, probably nothing. But if you have a deep in the money option or something that there's some value there.

Kelly: So, putting that question aside, the question is, can the Fed buy a portfolio? We think about what the corporate bond interventions or some market based intervention. Can the Fed act as dealer of last resort in derivatives? And the Fed has this interesting authority section 4.4 of the Federal Reserve Act, which include, its incidental authorities clause. So basically this is anything that the Fed needs to do to be a bank that's not expressly forbidden in the Federal Reserve Act that's, as long as it's still consistent with the Federal Reserve Act, the Fed can do.

It was just thinking about, how is the Fed going to take collateral in the commodities markets? And one of the things that's very difficult is making sure it's secured to their satisfaction, especially when things are so volatile. You mentioned oil going negative in April 2020. I talk about July to December 2008, which a barrel of oil goes from $145 bucks to $31. So, very hard for the Fed to be sufficiently secured in an environment like that. And a possible solution is that the Fed could take a derivatives position consistent with the so-called long position in commodities that they would be taking.

Kelly: So, if the Fed wants to serve popcorn in the lounge like my hometown bank did on Fridays, they can do that without Congress writing a law, right. And so they actually use section 4.4 when they set up SPVs. Because you're like, what the heck? The Fed’s being all entrepreneurial, it's starting businesses in the middle of a crisis. But it's incidental to their discounting authority. So it's this catch-all authority. And it's not meant to be used expansively. But they use it in 2008 when they are setting up the Maiden Lane facility. So just as quick backdrop, the Fed is helping JP Morgan rescue Bear Stearns. And JP Morgan says, look, there's 30 billion of assets we don't want. So they set up this SPV called Maiden Lane. JP Morgan takes the junior most billion of the portfolio. The Fed takes the senior 29 billion. And the Fed is effectively consolidating this on its balance sheet. It's much like a market based facility, where the Fed is just sort assuming these assets. And as part of this operation, the Fed basically takes a pro rata share of bears derivatives in this book that they're taking on. So there's a lot of macro hedges and credit hedges and things like this. And the Fed basically uses its section 4.4 Authority to think about, these are incidental, and then there was some value in the derivatives as well.

Beckworth: So, it never violated the mantra that you'd often hear Jay Powell say, “we lend, we don't spend.” It was able to honor that principle.

Kelly: Right. And so, part of why it uses section 4.4 is, section 13.3 mandates that the Fed can only discount notes, drafts, and bills of exchange. These are effectively instruments of credit. And you can think about derivatives adjacent to that. But it's not always clear, especially when there's not a lot of value, a lot of market value, if it's just a math to money option or something. But when you're taking a portfolio, when you're doing portfolio purchasing, it's incidental to that portfolio. You have a dealer who's got basically a match book, they're going to have a million derivatives. And you don't necessarily want to just take the long position and force all these derivatives to get shut down, have to liquidate all the derivatives. So that's kind of the distinction.

Beckworth: And for our listeners who don't know, SPV is special purpose vehicle, and it allows the Fed to set off an off balance sheet entity. Is that right, to engage? What's the role, I guess, of the SPV in the Fed? Why does it need to use it?

Kelly: It doesn't, is the short answer. It doesn't at all. And this is something that's commonly misconstrued. Basically there's this myth that has grown that, okay, the Fed does lending, there's this word, discount. So the Fed can lend, but if it wants to buy stuff, if it wants to buy 29 billion dollars off Bear Stearns's books, it has to set up an SPV. If it wants to buy commercial paper, it has to set up this new company, lend to the SPV and then the SPV can buy whatever it wants. And that is not true at all. The Fed has never said it does this. But I say the Fed, lots of Fed officials think this is why they do this. So you can read it in some of the books about 2008, they'll say, oh yeah, we did it to fit with 13.3, but that is not true from day one.

Kelly: Another shout out to YPFS. We have a great resource library of crisis documents. And included in that is a memo that came out in legislation, or in lawsuit against AIG, basically this first memo that the Fed wrote associated with Maiden Lane. And they are doing this analysis on a look through basis. They're saying we are discounting these assets. Discount is this old word, it means to buy paper from somebody else, right. So they look at it both ways. We can look at it as lending to a vehicle that's JP Morgan’s, or we can look at it as the consolidated Fed, basically discounting these assets. So, the SPV is totally for convenience. It's for accounting and transparency. And there's also a bit of legal protection. So, if Bear Stearns decides to sue the Fed they're suing Maiden Lane, as opposed to the whole Fed balance sheet. But yeah, it's not a legal gimmick at all. And the Fed has never said this. Like I said, they're very serious about this 13.3 authority, and they're not trying to play games with it.

There's this myth that has grown that, okay, the Fed does lending, there's this word, discount. So the Fed can lend, but if it wants to buy stuff, if it wants to buy 29 billion dollars off Bear Stearns's books, it has to set up an SPV. If it wants to buy commercial paper, it has to set up this new company, lend to the SPV and then the SPV can buy whatever it wants. And that is not true at all. The Fed has never said it does this.

Beckworth: So, I had Larry Ball on the show several years ago, and he had a really fascinating book about the Lehman part of the crisis, where the Fed did not step in. And of course, Lehman led to the run and the money markets and really intensified the financial crisis part of this whole period. And he argues the Fed could have acted differently, but it was a political decision. Whereas some Fed officials said, no, no, no, our hands were tied. So, you have a treasure trove of documents. You probably have thought about this, read about this. What is your take on that development?

Kelly: Yeah. So I don't want to be seen as too biased , because the beginning of this conversation, we talked about how Tim Geitner is cutting my paychecks. But I'm partial to Larry Ball in that, the Fed could have acted differently. But he gets stuck on this idea of solvency and a very narrow view of the Fed’s authority as being able to, as long as they can expect repayment. And a good idea he has is okay, what if there was a bridge loan? So maybe the Fed can still be secured to its satisfaction, even though then some long term creditors will take the hit instead. And Geitner is on record saying, yeah, we probably could have done that. But what we were hearing from traders is basically that this wasn't going to stop the run and we didn't think that was within our authority if we're not going to stop the run.

Kelly: So, maybe that's not my interpretation of 13.3, maybe it is. But that's the thing, it's like 13.3 you have a few pieces. You have to expect repayment. I mean the language of the law is the Fed has to be secured to its satisfaction. And that's a very discretionary, and there was much less guidance in 2008. And so they maybe thought it was appropriate. Like look, we shouldn't be lending to firms that are still going to fail. But there's a political piece too, of maybe you don't want the headline of “Fed lends to Lehman and then it fails five days later.” Even if that was their plan from the outset. Like, oh, we're going to give it a bridge loan. We're going to get our money back, but it's still going to fail. It's just going to fail more slowly. So I'm partial to Larry Ball in that, obviously it's hard to look back and say, yeah, letting Lehman go was great. And maybe there was collateral for some liquidity. But there's also this, you can't just look at the balance sheet. There's this ephemeral, the way traders behave overlay that we think about a lot here at YPFS, and of this practitioner's view of, you can have a pristine balance sheet with 10% capital, but if all of a sudden traders say it's not enough, then it's not enough. And you can have a balance sheet with 1% capital. And if traders like it, they like it. That’s a hard thing to balance.

Beckworth: So it's the expected balance sheet. And they thought the balance sheet was going to look really bad very soon, had they followed through and thrown this money at Lehman? Well, it's an interesting discussion and we'll provide a link to the show that we had with Larry Ball on this. So take a look. I just had to bring it up since you seem to know a lot about the 2008 crisis and your program there spends time on it. All right, well, let's move on then from 13.3 facilities and let's move to another area you've written on that relates to financial crises, or potential financial crises. And this is more actually in your toolkit of preventative measures. So we've talked a lot about responding after the fact of crisis. You've written about the standing repo facility. This has been something I've been following for a while too. And you have made some suggestions on how to make it better, more effective. So maybe walk us through that article. What did you try to show with the standing repo facility?

Improving the Standing Repo Facility

Kelly: Yeah, so first I'll say we know that the SRF, the standing repo facility isn't a panacea, right? I mean, at some point you made, it operates through its 35 counterparties. Some of which have the same parent company, it's like the bank and the dealer from the same parent company. And it depends on them being able to pass balance sheet to the rest of the financial system, if that's where treasuries are being liquidated. So we won't get this thing to be perfect. Like repos were great in 2019, responding to the repo blow up there. They were great in 2020 as the Treasury market started melting down. But the Fed is still coming in and buying 500 billion dollars of treasuries in two weeks in late March 2020. So, this is really at the margin. But the way I was trying to think about this is of with respect to 2020, which what we saw was really a dash for cash. The SRF is great at replacing lost repo funding. And that's kind of how we think about runs sometimes, as you hit a liquidity air pocket and you can no longer fund your positions. But we've also built up this financial system, especially post 2008, that depends on the liquidity of treasuries. These are your high quality assets. These are the things that can get you liquid in times of instability. But the basic nature of financial instability and crises is that there are too many people liquidating things at the same time and there's not enough balance sheet to take it on. So, there's that tension.

Kelly: And basically, what I talk about with the SRF is that, okay, the SRF is great at replacing this liability side, but if the asset side blows up, as treasuries can do in a given day, then we're eroding capital and the repo facility is only going to lend against the new value of the Treasury. So if everybody's liquidating treasuries and they go down four or 5%, suddenly your repo funding goes down by four or 5% too. And whether you get it from the market or you get it from the Fed, they're going to do that same mark to market. So this kind of goes back to this idea of the derivatives authority. And I sort talk about, well, if the Fed can take a match book position, again, take a portfolio view. There's more the Fed can do because it's more secured. It has a futures contract associated with the Treasury that hasn't lost value because they're much more liquid than a cash treasury, which has to be fully funded would be.

Beckworth: Okay. Well couldn't one of the roles the standing repo facility plays is just expectation management? If it were credible, if it were fully functional and all that, you may never get to the point where you need to use it. I mean, that's one of the ideas. And so that's my hope, that would minimize the need to go through all the hoops that you just mentioned. But I guess one of the concerns I see with it is just the amount of participation in it so far.

We've also built up this financial system, especially post 2008, that depends on the liquidity of treasuries. [...] The basic nature of financial instability and crises is that there are too many people liquidating things at the same time and there's not enough balance sheet to take it on. So, there's that tension. [...] The SRF is great at replacing this liability side, but if the asset side blows up, as treasuries can do in a given day, then we're eroding capital and the repo facility is only going to lend against the new value of the Treasury.

Beckworth: And I was at a conference where they mentioned this and some regional banks didn't want to get into it. Just there's all these regulatory requirements and the way it affects their balance sheets, it was a real hindrance. And you really need to push it from the top. The vice chair for supervision at the Fed and others beneath him need to really push this as something that needs to be a regular part of your daily business. Uses facilities in normal times so that in stressful times, it's just routine. Otherwise it's going to become another discount window that's not you. So, do you get a sense that's still a concern? Are we seeing more pickup and counterparty involvement in the standing repo facility? Or does more need to be done from the top to encourage its use?

Kelly: Yeah, it's tricky. I mean, so far I think we've been lucky that the SRF, for whatever reason, it just doesn't have the stigma of the discount window. Even though you could take those same treasuries to the discount window and get the same amount of money at effectively the same rate, it just doesn't. And also the disclosures of you borrowing would happen on the same timeframe. But for whatever reason, it doesn't have that stigma. And I think that's good. I mean, again, where the Fed basically underwrites the whole Treasury market, at least from volatility, right. I mean, it's decided for good reason, it doesn't want to do yield curve control. It doesn't want to do things like that. But there's an incentive for the market to be able to implement monetary policy. We talk about the Fed's balance sheet, if the SRF can be effective, the Fed's balance sheet can get a lot smaller in this world. Because it has this contingent financing facility that can keep people in Treasury positions, can help the market basically implement the long end. And so we'll see. This goes back a little bit to Lev Menand too, where you start opening this thing wider, and then it's like, okay, well it's come one come all at that point. And it's just an open market.

Beckworth: Right. That was his comment.

Kelly: Is this an open market operation? You want this to be able to reach every balance sheet. But also you're bound by the law. So it gets tricky at this point of, whose it for, and then does it solve the problem? And it might not be able to solve both of the things at once.

Beckworth: Yeah. He was worried that if you made it too accessible, that you just encouraged more shadow banking, and more use. But, I think it can play a role if managed wisely, and maybe we'd have to just kind of figure it out as we go along that path. Okay. Well, we're getting near the end of the show. But before we go, I want to spend some time with you on stablecoins. So, it's been a hot topic lately and also related is the discussion of central bank digital currency. The Fed had a paper come out, not too long ago, where it endorsed wholesale CBDCs, but was completely against retail. And you could maybe foresee a world where wholesale CBDC worked in hand with a stablecoin. But what are your concerns with stablecoins? And do they fill a niche, or do you think they're taking up space that could be provided by another financial service?

Stablecoins

Kelly: Yeah. So my concern with stablecoins is really this emerging consensus, which is almost bipartisan in a way, with this idea of, okay, we’ll let anyone start a stablecoin if they want, but they have to back it with really safe stuff, and we have to be sure they are doing that. So either that's market transparency where they have to provide disclosures, or we send in a supervisor and they say, yep, this is all treasuries and Treasury backed repos backing this. And that seems good. And it's good for the holder of the stablecoin. But there's potentially risks to the traditional financial system, which underwrites 100% of real economic activity, lives in the traditional financial system as of today.

Kelly: So there's kind of a few steps to this. So a year ago, this stablecoin story was, they're holding all this commercial paper, that's risky. We can all agree on that. We see MMFs get bailed out, money market funds get bailed out because they're holding commercial paper, fine. So we want to shift them into safe assets and make them look more like a government money market fund, which runs in during crisis often. So okay, that's fine. But there's a couple pieces to this. One is you still have a new intermediary. So it's still one step in between what was conceivably retail deposits or things like that and the financial system. So you've reconfigured bank liabilities. And when we talk about MMF runs, we say like investor runs, but sometimes it's the intermediary itself that runs. So there were examples in 2008 where the intermediary had rock solid Treasury collateral, or MBS collateral. And they're like, you know what? We're getting phone calls because people are worried we're exposed to Lehman. And yes, we have Treasury collateral, but we don't want to deal with it. We don't want the speculation. We don't want to tell people that we have a repo exposure, so we're just going to dump it. And then Lehman's in trouble. So that's the kind of thing where having another intermediary is not good.

My concern with stablecoins is really this emerging consensus, which is almost bipartisan in a way, with this idea of, okay, we’ll let anyone start a stablecoin if they want, but they have to back it with really safe stuff, and we have to be sure they are doing that. So either that's market transparency where they have to provide disclosures, or we send in a supervisor and they say, yep, this is all treasuries and Treasury backed repos backing this. And that seems good. And it's good for the holder of the stablecoin. But there's potentially risks to the traditional financial system.

Kelly: But the bigger piece here is, because these are non-bank stablecoins and they live on the blockchain, and they're primarily used to trade in cryptocurrencies. If you have a crypto selloff, which tends to be procyclical, it tends to happen at the same time that bad things are going on in the traditional financial system. And you want to get your stablecoins. You say, forget the crypto economy, I'm done with it. I got burned on Ethereum and Bitcoin. I want my deposits back. You have to go to Tether or Circle and redeem. And even if they can meet those redemptions, which is what the emergency consensus is all about, it's, can you meet consumers redemptions a dollar for dollar? Okay, great. Even if that happens, you still effectively have a massive money market fund liquidating into what is presumably a volatile environment. And it's worse that it's liquidating Treasury backed repos than it would be if the fund held Bitcoin and Tesla shares. Because it could liquidate all that and we'd all be fine. And if it starts liquidating like a money market fund, then there's more macro trouble.

Beckworth: Well isn't this just an expansion of financial innovation. I mean, financial industry's always facing up against new innovations, new products, new ways of doing things. We didn't have crypto before. There may be ongoing demand for safe assets as we get to the other side of the pandemic and aging populations and such. I mean, is it really surprising, I guess that we're having this as a new financial product?

Kelly: I wouldn't say surprising. But it's also the question of, what is the purpose? So, we've been talking about shadow banking risks, and yes, there's a ton in money market funds or things like that. But at least there's an obvious demand side problem that's being solved. What's in stablecoins is not corporate treasurers with too much cash and nowhere to go with it, right. So we're adding risk to the system. And the other thing is that they really gobble up safe collateral when we do this. And that adds to macro risks as well. We're coordinating off safe collateral in these things as we let them get bigger. And the innovation can happen. I mean, you see banks doing this already. JP Morgan is really leading in this space of tokenizing bank deposits. So, you basically have, it's still a bank and it's still a deposit, and it can live on the blockchain. And that's probably the future here. So yes, the innovation is great. But there's no reason we have to let it infect the traditional financial system, especially money markets. We talk about innovation and the way it bears out as risk in finance. You think about, okay, the dot com bubble. That stuff's all well and good. It's like, there's no reason to take it straight to the money markets on day one. And as soon as something goes wrong in Bitcoin, people are bailing on stablecoins and all the money market assets get dumped onto the system.

Beckworth: Well, you answered my question. But just to elaborate on it. The future of stablecoins you see is being done within the traditional banking system, as opposed to being a bunch of standalone entities?

Kelly: I hope so. I hope so. Like I said, emerging consensus is a little different. But to me that seems the safest route with the least chance of contagion and infection. And just to your early question on CBDC. I don't know. I hesitate to take too much of a view on this, because it's weirdly become a partisan issue. And there's an interesting alternate universe in which there's like no Bitcoin and no crypto culture. And it's just the Fed saying, hey, we're thinking about using distributed ledger technology to improve our transfers abroad. I mean, is that even a press release? I mean, nobody in Congress cares about that at that point, right. So I don't know. There's a lot of talk that FedNow and the instant payments will deprioritize this a little bit. I'm not super convinced. We talk about tokenizing deposits. There's a big matzoh ball to be had and tokenizing securities in general and having real assets on the blockchain. So I could deliver a case of hot dogs to you for your labor day cookout. And we scan our phones and I get money from you in real time. And you take acceptance of the goods. I mean, I don't see that happening on FedNow. And it just doesn't have those rails. So we'll see it. There's a lot of innovation happening in banks. And that could be, there's going to be a lot of interaction between the banks and central banks at this point.

Beckworth: Okay. With that our time is up. Our guest today's been Steven Kelly. Steven, thank you so much for coming on the show.

Kelly: Thanks David.

Photo by Chip Somodevilla via Getty Images