Steven Kelly on the Challenges of Treasury Equity Funding for 13(3) Facilities

Steven Kelly is the associate director of research as the Yale Program on Financial Stability. Steven returns to the show to discuss his new model, the Treasury Equity Model of the Federal Reserve’s emergency lending. 

Check out our new AI chatbot: the Macro Musebot!

Read the full episode transcript:

This episode was recorded on January 23rd, 2025

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: Welcome to Macro Musings, where each week we pull back the curtain and take a closer look at the most important macroeconomic issues of the past, present, and future. I am your host, David Beckworth, a senior research fellow with the Mercatus Center at George Mason University, and I’m glad you decided to join us. 

Our guest today is Steven Kelly. Steven is the associate director of research at the Yale Program on Financial Stability. Steven joins us today to discuss the Treasury-Equity Model of Federal Reserve emergency lending. Steve, welcome back to the program.

Steven Kelly: Great to be here.

Beckworth: This is number five, I understand, but the first time in studio, in person.

Kelly: That’s right. I came for the five-timers jacket.

The Discount Window and Its Current Momentum

Beckworth: That’s right. It’s great to have you back. You’re a regular on the show. Last time we were together, we were talking about a topic near and dear to your heart, my heart, and some of our other guests, and that’s the discount window, how to make it better used, more efficient, maybe minimize some of the stigma. It was a great conversation, and we talked about the developments in the banking regulatory front from the comptroller of the currency to Michael Barr. They’re all pushing for more regular use of the discount window.

Bill Nelson’s been on the program. He’s had some suggestions. You had some suggestions, better reporting measures or practices, aggregating the data up so there’s less ability to figure out who’s using it. Senator Mark Warner also had pushed a bill through. There was a lot of momentum last time we were together for the discount window, and I think we both think this is a good thing.

My question is, where is that momentum now, where you have a new administration? Even the culture of the Fed seems to be changing a little bit with this new government. Where do you think this is going?

Kelly: At the least, I think it’s on pause. Things like this are always in motion at the Fed and elsewhere, but we don’t have a vice chair for supervision at the time of recording this. We’re going to get a new comptroller. I think what may change, to the extent this moves forward, is the way the previous crop of regulators really thought about liquidity things was as an add-on.

Michael Hsu had his ideas for how to add on a requirement for liquidity that was shorter term but allowed credit for preposition collateral. Michael Barr had his proposal that now dates back to May of last year, which ended up going nowhere for requiring some prepositioning. I think if the Trump crop of regulators, if we got a full crop and they were picking it up today, there would be more appetite for incorporating the window into existing liquidity regulations—into the LCR, into internal liquidity stress tests, and giving credit to banks for collateral that they have at the window.

It’s something like $3 trillion of collateral at the window. Right now, as far as liquidity regulations are concerned, that’s valued at zero. There would be, I think, some carrot with the stick of prepositioning requirements or things like that.

Beckworth: That’s interesting. If I recall correctly, I believe that was Travis Hill’s suggestion when I had him on the podcast.

Kelly: Miki Bowman, who is shortlisted, maybe by the time this podcast is out, will be the vice chair for supervision, or in process, has said similar things as well.

Evaluating the Treasury-Equity Model of Fed Emergency Lending

Beckworth: There’s some interest still in this. Maybe we’ll see progress by the time this show comes out, if not later. You’re here today to discuss a new paper, a really interesting paper. I enjoyed reading it. I learned from it. It is titled, “Of Last Resort: Evaluating the Treasury-Equity Model of Federal Reserve Emergency Lending.” Anybody who loves talking about crisis, Fed facilities, past few years, actually going back to 2007, will enjoy this paper—the history of the Fed. This is your job. You get paid to think about these issues. Maybe give us the executive summary of it, and we’ll go down section by section in the paper.

Kelly: Yes, sure. It’s a short history of what I’ve ended up calling the Treasury-Equity Model of Fed emergency lending. Those who write papers know how these expressions come together. I did not make up Treasury-Equity Model. I had it in brackets as I wrote the paper, expecting to come up with a better name, and never did. I just took off the brackets, and now it’s called the Treasury-Equity Model.

What it describes is the increasingly common way the Fed does emergency lending, in which it takes a junior tranche, so to speak, of funds from the Treasury as first loss absorption. What the paper does is really evaluate the reasons for this, historically speaking, as well as look at the increasing use of this method, and what the payoffs have been. It’s a short history. It only goes back to 2008, like most things with the Fed’s 13(3) emergency lending authority and facilities. Obviously, there’s a longer history, but it really got interesting in 2008.

There’s a few things the paper evaluates. One is there is this legal origin for why the Fed thought it might be necessary. There’s an old memo that the Financial Crisis Inquiry Commission released to the public that shows the Fed is footnoting. Here’s why we think some equity might be helpful. We’re not sure. That’s basically the last we have in the public sphere on the legal reasons, and then it became like this financial weapon.

Folks may remember from COVID, the idea was like, “Oh, we’re going to leverage Treasury money. Treasury is going to be our equity investor, and we’re going to leverage it into a bunch of Fed lending that we couldn’t otherwise do.” Powell and Secretary Mnuchin came out and said, “Oh, we’re going to do up to $4 trillion of lending through the Fed. We’re supercharging Fed lending.”

I think the idea or the hope is that there’s some Treasury spend to go with the Fed’s ability to lend. What I show is that facilities in which the Treasury has invested, it ends up push[ing] around the Fed quite a bit. The outcome of these facilities, even before, not judging when the facility is over, but even ex-ante expectations, is a facility that’s stringent enough that the Fed doesn’t need the Treasury money.

The Fed is being consistently held back by the politics of the Treasury. What I suggest is the Fed should rethink this relationship. It’s not as though every time the Fed does an emergency lending facility, the Treasury says, “Hey, we want some of that action. We want to give you some funds. We want to go to Congress and ask for money.” It’s really the other way around. The Fed should rethink how trigger happy it is to go to Treasury for this money.

Beckworth: This is a great illustration in my mind of how when the Fed does veer into fiscal policy space and it entangles the fiscal policy via the Treasury-Equity here, it always raises questions of its independence and its ability to do what it wants. I know in crisis it has to do things differently, be more aggressive, take risks. It’s a great illustration because if the Fed’s abilities to perform in this sphere are undermined or questioned or politicized, it could easily bleed over into, say, monetary policy.

It’s a great way to think about this issue, and it was a great history of it. Again, it’s on how best to use these 13(3) facilities. Maybe walk us through what is the 13(3) facilities, its history, what does it allow the Fed to do, what can the Fed not do, and how does that shape the discussion?

The Fed’s Section 13(3) Facilities

Kelly: Sure. The Fed’s Section 13(3) facilities, it’s Section 13(3) of the Federal Reserve Act, as most listeners will know if you listen to the past episodes, allows the Fed to lend beyond its typical discount window authority. The Fed has monetary policy, it has a discount window, and it has 13(3), which historically has meant lending to nonbanks. We already have the discount window for banks. There are some exceptions where 13(3) facilities are designed for banks.

Under this section, a supermajority of the Fed board must authorize that there are “unusual and exigent circumstances.” When the Fed does liquidity provision under this statute, it has to be “secured to its satisfaction.” What it has historically interpreted that to mean is that it expects repayment in most states of the world. If I loan you $100 and expect $80 back, that’s basically a $20 grant, effectively fiscal policy.

The Fed has to be a prudent lender, ex ante, and expect to get paid back. That’s the starting point. It has been both lending facilities and purchase facilities, across the broad sweep of markets in 2020 in particular. That’s the authority, that’s what it’s there for. It was originated in the Great Depression and basically quiet until 2008, then it was invoked many times in 2008, many times in 2020, and again after SVB failed in 2023.

Beckworth: The wording, and this in your paper in Section 13(3), says it has to be unusual and exigent circumstances. That’s defined by the Fed, I guess they have the power to define that. Obviously 2008 would be an example. 2020, I don’t think there’s any debate about that. Maybe there might be some other times where maybe we can raise questions about. I guess since Dodd–Frank, it also has to be applied to more than just one firm.

Kelly: Now it has to be a broad-based facility. The idea was the Fed rescued AIG, 2008 the Fed rescued Bear Stearns, or contributed to the rescue of Bear Stearns, and Congress said, “No more. From now on if you’re going to do a 13(3) facility, it’s got to be broad-based.” Which the Fed itself has defined to mean at least five eligible borrowers. Think about it as going after certain markets or certain sectors of financial markets. That’s what the Fed can now backstop.

Beckworth: You have a really great table in your paper where you show all the facilities that were used. As I was looking at this, I believe this is figure 1 in the paper. I’m looking at it now for those who are listening. You have the programs from the Great Financial Crisis of 2007 and 2009, and then COVID-19 pandemic, and then most recently the banking crisis of 2023, the BTFP, there’s only one facility then.

What’s interesting as you look at this—and you make a lot of points, a lot of data on this chart. In fact, I had to blow this up for my old eyes to read it. It’s striking if you look at the facilities for the Global Financial Crisis. A lot of them are directed at single firms. You even see it in the names: Bayer, Maiden Lane, AIG, revolving credit facility. You go down to COVID-19, it’s the more general, money market, primary dealer, or the corporate bond—it’s definitely more than one firm. You can clearly see the history of Dodd–Frank reflected in this table.

Kelly: There were programs like that in 2008 as well. The commercial paper facility, a facility for primary dealers. The germ of them was there. It’s those facilities that Congress effectively endorsed and said, “We don’t like you doing these ad hoc things to Citi or whomever.”

Beckworth: What’s the motivation? You don’t want to pick a favorite winner, loser?

Kelly: I think that’s a big piece of it. On top of the fact that it must be broad-based, Congress also added it cannot be for saving one firm from bankruptcy or to prevent a failing company. They said it like three or four times in the new law: We don’t want you picking these winners. It makes sense in the politics of the moment and how mad people were about Lehman Brothers and AIG and all of that.

Beckworth: It also says now in Section 13(3) that you got to have the approval of the Treasury secretary, which was new with the Dodd–Frank changes to the act. As you note in your paper, that was already done.

Kelly: The Fed got sign-off from the Hank Paulson Treasury in every case in 2008. Part of the fun history of that time as well is that there was no official Treasury sign-off as there is now. But in many cases, the Fed got what Hank Paulson would call his “all money is green” letters, which were these public letters that he would write to Tim Geithner or Ben Bernanke saying, “Hey, we support this facility that you’ve just done for Bear Stearns or for AIG, and we understand that it may reduce the earnings that you remit to the Treasury if you have losses. That’s just a fact.”

There is no indemnity from Treasury there. We could talk about the challenges of getting Treasury indemnity, but it was meant to show public political support from the Treasury Department. It was this all money is green, just stated fiscal facts, but it was meant to show Treasury support for a Fed facility.

Creation of Section 13(3)

Beckworth: You mentioned that 13(3) was created in the Great Depression, which makes sense. What was this motivation then? Who used it? What was it for?

Kelly: Effectively nobody used it. It quickly got superseded by additional statutes that were added temporarily. There was $1.5 million—that’s a million with an “M”—of lending done under 13(3) in the Great Depression. I’m not sure what that is in today’s adjusted dollars, but it can’t be much. Then the authority went dormant. It was invoked but not lent under to help with back when it mattered to be a nonmember of the Fed. It was actually for bank-type institutions in 1966 and 1969, and I think for one bank in 1980. Again, no loans in any of those cases. Obviously in 2008 it gets used many times.

Beckworth: That’s interesting. Thinking back to the Great Depression, we didn’t have a shadow banking system like we do today. Maybe there’s some dollars on the margin outside the regulatory—

Kelly: Yes, the loans went to typewriter manufacturers and things like that.

Beckworth: It’s interesting because today 13(3) would be used for more of that space where you do have shadow banking, dollar creation outside the regulatory—

Kelly: Typically, exactly.

Beckworth: You can see an evolution of its importance or its emphasis would change as the financial system has changed.

Kelly: That’s right, and I think that’s at least a partial explanation for why it seems like we’re invoking it so frequently in modern history.

Beckworth: Let me park there just for a minute because one of the things I’ve thought about is the Fed’s balance sheet is always going to be tapped when we have these really big financial crisis moments. If for no other reason that the world is dollarized, we have a global financial dollar system, and the Fed in some sense has to respond. It has to open these facilities up. It can’t just say, “Oh, that’s euro-dollar market. We’re not going to worry about that,” because it would effectively come back home and bite us in the rear. It seems to me, at least on some level, it makes sense you’re seeing more use of 13(3).

Kelly: I think that’s right. If you look at global data, not just the US, financial crisis frequency and financial crisis intervention frequency is rising over the last several decades, post-Bretton Woods, and so is net growth. It feels like a paradox that the optimal level of financial crisis is not zero for optimal growth. We’ve seen those things rise in concert, and it makes sense that 13(3) would follow that.

Beckworth: Let me throw out a hot take I have on this. You may have heard it on the podcast before. Some people might question the use of 13(3). They might say there’s moral hazard or you’re just incentivizing more dollar creation outside the regulatory fence. I say, “Hey, we’re creating more demand for dollars around the world.” So stronger demand for dollars, more seigniorage flows into the US. We can sustain our budget deficits longer. Maybe there’s a silver lining in the fact that we do use 13(3) that we don’t think about. It’s a point. It’s an implication that flows out of it. Any thoughts about that?

Kelly: I would say it’s generally a good thing for America that there is no credible commitment to allowing the financial system to collapse. Moral hazard should be managed in its own right, but I don’t think there should be a risk of, hey, it’s the pandemic. Maybe the Fed’s going to sit on its hands. That’s not a world we want to live in either.

Beckworth: I guess my point is, as a side benefit, we’re extending the reach of the dollar by doing this, which in turn is great for America.

Kelly: Yes, and the swap lines and FEBA repo are arguably bigger contributors to that as well.

Beckworth: Good point.

Kelly: At least the overseas dollar market.

Beginnings of the Treasury-Equity Model

When the Fed is doing this, this is the first time they say, “Hey, it would be great if we could get an indemnity from Treasury on this.” Hank Paulson says, “Yes, we’ll do whatever we can.” Turns out the lawyers don’t like it, and so what does the Fed do? They go, “We expect to be repaid. We feel like we’ve met our secure to the satisfaction standard, so we’ll do it.” You can see that the preference from the Fed was, oh, it would be nice to get some Treasury protection here. Do we need it legally or financially? We don’t think so.

That’s step one. Come the fall of 2008, the Fed is setting up the commercial paper funding facility, and this facility buys both asset-backed commercial paper and unsecured commercial paper. Folks may remember that asset-backed commercial paper, ABCP, was one of the financial instruments at the heart of the financial crisis. There was a lot of subprime paper, a lot of securitized finance, which was being dumped across the board. The Fed was fine buying this, supporting this market, the highly rated tranches of it.

It said, “Look, this is collateralized lending per usual. We’re going to buy commercial paper,” but there’s all these assets in the vehicle that issues the commercial paper, whether it’s housing or whatever else. It’s secured lending in a traditional sense. We also want to buy unsecured paper. This is the payroll financing for General Electric and McDonald’s and Coca-Cola. They want to support that market.

These are highly rated companies that have a balance sheet behind them, but the Fed is legally uncomfortable with their ability to buy something that’s wholly unsecured. If the Fed goes out and just buys unsecured paper from General Electric, this is what I was talking about, they have this memo from the time where they go, “Ah, it’s not clear we can buy something that’s wholly unsecured. We should probably get some collateral.” They intend to get $10 billion from the newly legislated TARP funds, which Congress has just passed.

Hank Paulson in the Paulson Treasury, again, approves of everything the Fed does, loves what the Fed’s doing, supportive of Ben Bernanke and Geithner, and says, “Sorry, we’re not going to give you the $10 billion because I don’t want this to be the first TARP program.” That’s clearly a political consideration. You can read Ben Bernanke’s memoirs, and he goes, “To this day, I’m vexed by this.”

What the Fed did is it said, “Okay.” They write this memo, and they decide that a good way to get some collateral behind the unsecured paper is we’ll just charge an insurance premium to the unsecured issuers. We’ll charge them an extra 100 basis points, and we’ll pool that, and that will be our pool of insurance premiums that are available to pay out losses against any one issuer of commercial paper.

If you’re keeping track, they’ve gone from 0% secured to 1% secured. It makes sense on a financial basis because the unsecured commercial paper is actually much safer. The Fed officials have said this, like, “This was totally a legal thing. The unsecured paper is actually much safer because you have a whole company behind it. You have a balance sheet that you have claims to as opposed to a shell of securitized assets.” In the case of ABCP, there is no company that issues it. It’s just a securitization vehicle.

Anyway, that’s how they got around it. This pooling concept has become central to the Fed’s facilities now that Dodd–Frank requires everything be a portfolio for 13(3). It has to be broad-based. Part of what I argue is this risk management feature is now built into 13(3) by law. The situations in which the Fed used to think, “Oh, we need 1% collateralization from this extra surcharge,” it now gets by law because it has this pooling feature.

Beckworth: That is so interesting. In the case of the funds sought for Bear Stearns, Treasury says no. You mentioned Paulson didn’t want it to be the first use of funds for TARP.

Kelly: That was for the commercial paper facility.

Beckworth: Commercial paper facility.

Kelly: Sorry, I should have clarified. At the time of Bear Stearns, there’s been no fiscal allocation yet. The only money that Treasury has is in the ESF at that point, the exchange stabilization fund. There’s $50 billion in there. Paulson says, “I’ll do what I can.” His lawyers say, “Bear Stearns doesn’t meet the requirements for exchange stabilization fund usage.” 

They get the same rejection when they go to rescue AIG. Again, they want the indemnity. Lawyers say no. The ESF in its entirety gets used as a guarantee fund for  money market funds. The lawyers like that more presumably because it’s more systemic. You have a multitrillion-dollar market as opposed to an individual firm. Obviously we’ve seen the ESF used tremendously since then.

Beckworth: It just seems like Paulson said no quite a bit during this initial run.

Kelly: Again, he didn’t want to except for—and this is the point I make. Paulson loved everything, signed off, wrote these “all money is green” letters to the Fed constantly. They all got along great. As soon as Bernanke says, “Hey, we just need $10 billion for the CPFF so we can buy unsecured paper.” Paulson says, “I don’t like the politics of this being our first TARP program. We’re going to say no.” That’s where it breaks down and the Fed has to find other ways to secure itself.

Beckworth: The point I think you’re making is even though the Fed did not get the funds from Treasury, it was just the desire and you see the politics associated with getting those funds, whether you get them or not.

Kelly: Yes, and then the Fed went and charged these issuers an extra 100 basis points. GE and others had to pay the surcharge. If you fast forward to 2020, the Fed rolls out the CPFF too, gets $10 billion from Treasury in this case, and there’s no surcharge. The borrower is better off not having to pay that surcharge. That was the cost of Paulson balking at that point.

Beckworth: You also note that he did say yes for the term asset-backed securities loan facility.

Kelly: Yes. This was announced in November 2008 and really effective in 2009. This was the first actual effective usage of the Treasury-Equity Model. The Treasury had these TARP funds and invested $20 billion of junior funding into what was going to be a $200 billion Fed lending facility against securitized assets. You have a 10 to one leverage ratio inherent in that. It seemed to work out okay. There weren’t a lot of politics associated with the unwind, which ended up not being the case in 2020.

There’s some reporting done on the facility in real time where Bernanke is saying, and Bill Dudley and others are saying, “Hey, even if the economy trends much worse than we expect, we expect no losses for the Treasury or the Fed.” If you remember back to our conversation on secure to the satisfaction, that’s the standard is the Fed is secure to the satisfaction if and when it expects no losses. No losses and expectation.

It shouldn’t be probability zero of loss, because then you’re never going to be a lender of last resort. It’s not that they can’t ever take losses. It’s that you can’t ex ante expect a loss. If the Fed is already in that world, as it was in the TALF, because it said, “We don’t expect any losses for Treasury or for us,” you don’t need the junior funding from Treasury. In this case, it seems like it didn’t really matter. It’s not like we ran out of TARP funds, or Treasury really hamstrung the program. It mattered for the CPFF, and it would matter in 2020.

Beckworth: Are you suggesting then that the experience around the Great Financial Crisis indicates that the Fed could have met its secure to satisfaction need without going to Treasury?

Kelly: Yes, I think that’s right. It was a legal reason for the CPFF. Like, “Oh, we don’t think we can do this unsecured thing.” I think that has been put to bed by Dodd–Frank and this memo by the legal division on this pooling mechanism they used. There’s the financial version, which could work. If the Fed came out and said, “Hey, we expect $10 billion of losses on this, and Treasury is going to backstop those,” or “Congress allocated some money and they want to invest in Fed facilities, so we’re going to spend down the Treasury money,” that’s adding risk tolerance to the Fed that it couldn’t otherwise take. That’s effectively not what happened, because as they said in real time, we expect no losses even for the Treasury.

Treasury-Equity Model During COVID-19 Pandemic

Beckworth: Okay. Well, let’s move forward then to the next manifestation of this Treasury-Equity Model, and that’s COVID-19 pandemic.

Kelly: Again, we touched on some of this already, that when COVID started, there was $95 billion in the ESF. The Fed used some of that money to get rid of the surcharge in the CPFF. There was a similar facility for money market funds that bought some unsecured paper now because it had $10 billion. Then Congress really co-signs it. In the CARES Act in March 2020, they put $454 billion into the ESF for the Treasury secretary to support Fed programs.

This is when we get the corporate bond facilities, the municipal liquidity facility, the mainstream lending program. The money goes into all these programs, and this is when we get the big pronouncements about $4 trillion, the Fed’s going to lever this Treasury money into $4 trillion of Fed lending. I think the understanding, and even the reporting around at the time is, oh, this is for the Fed to take losses.

Again, the Fed isn’t supposed to never take losses. It’s supposed to not expect losses when it signs a program, avoiding fiscal policy, avoiding grant making. That really was the idea, I think, is that this is here to enable us to do lending that we couldn’t do absent the Treasury money.

Beckworth: Do you think the $450 billion that was allocated—that was in the CARES Act, right?

Kelly: That was in the CARES Act.

Beckworth: Was that allocated because people were like, “Hey, remember back in 2007, 2009, we went to Treasury and it almost worked. In one case, it did work. Let’s go back to that model.” Were they falling back on the old playbook?

Kelly: Yes, partially that, and partially the Fed has a lot of competence in lending and getting money out to various folks. This was a way to add some fiscal firepower to these Fed programs. In the event—and Nick Timiraos and Jeanna Smialek, past guests, have done really great reporting on this—you really see the Treasury being the one that’s holding the Fed back from being as generous as it would like to be.

Again, and this paper that I write says nothing about, oh, the Fed should be more generous or the Fed should invoke 13(3) more or less. This is taking all the Fed decisions as given, all the Fed desires as given, all their assessment of secure to the satisfaction as given. I guess I’m trying to play therapist and say, “You’re in an abusive relationship with the Treasury based on this history. I’m seeing these patterns and you’re not married yet. There’s nothing in the law that says you must take this Treasury money. It’s time to demand some changes from Treasury or to get out of this.” 

Because really what we see is that either the Fed and Treasury put together a facility where both expected no losses, which again, the Fed can do by itself, or the Treasury actively held the Fed back from doing something more generous. For instance, the Congressional Budget Office scored the fiscal impact of these facilities as having a $0 impact on the budget. That tells you right there, again, ex ante, the Treasury is expecting no losses. The risk tolerance just isn’t high enough to be accepting Treasury money.

Beckworth: It was really surprising for me to read in your paper how secretary of the Treasury, Mnuchin, he in some sense micromanaged some of these facilities. I imagine he turned some responsibility over the Fed, but if he saw something he didn’t like, he’s like, “No, and you’re going to go no further than this line.”

Kelly: Right. Nick has a line in his book about how the Fed governors disagreed about exactly how generous these facilities should be. They never had to solve their disagreements because all of them were effectively more generous than Mnuchin. Mnuchin was very concerned about the politics, very concerned about going in front of Congress. As I’ve presented this paper in various fora, some of the thought is like, “Well, it is nice to have the Treasury secretary through the Fed. It’s real nice to have the Treasury secretary in between you and Congress.” That’s totally true, but that doesn’t mean you’re going to be able to fulfill your mandate as lender of last resort.

Beckworth: It comes with a price.

Kelly: Right, the Fed would always love if J.P. Morgan handled emergency lending and the Fed didn’t have to do it, but J.P. Morgan’s not going to come in with the systemic mandate that the Fed has. That’s what we see in these facilities. 2020 obviously gives us the most examples where the Fed thought its statutory mandate suggested it should be more generous with these facilities and it could, within the law, within being secure to the satisfaction, and the Treasury was holding it back because of the politics.

Beckworth: What was Mnuchin worried about, that he was bailing out things, there’d be an image that he’s being too generous?

Kelly: Yes, he was worried that this was somehow backdoor refinancing, that he was getting snookered, that particularly politically divisive sectors of the financial markets were going to take advantage of Treasury. He was really influenced by TARP in 2008, which was a very different program. It was equity investments in banks and the Treasury had a net gain on those programs and he really wanted to show that too. Again, the Fed doesn’t need that money if he’s expecting to turn a profit.

Beckworth: I find it useful to think about the facilities during the 2020 period as there’s liquidity facilities and credit facilities. I’m invoking Lev Menand’s classification, you may disagree with it, but there’s some of them that are clearly like things that we did in 2008, the money market facility, commercial paper, primary dealer facility, currency swap lines—although that’s not 13(3)—but all of those things we do to keep the global dollar system running.

The ones that were innovative, new, and where you might say it’s more credit allocation, like the Main Street Lending Program. I can understand why he might get uncomfortable with that, but he was also uncomfortable with the municipal liquidity facility.

Kelly: That’s right.

Beckworth: Isn’t that in the Federal Reserve Act or some version of that where you can buy municipal securities?

Kelly: Yes, so the Fed can buy under open market operations munis up to six months; it doesn’t. This facility was designed to buy longer-duration municipal securities. It was a weird juxtaposition in that munis, they rarely default, they have corporate bond ratings, the same as corporate bonds, they have credit ratings, and  they default less even on a side-by-side basis.

The facility for them was less generous than it was for corporate bonds and it was precisely a political decision. We don’t want to get into muni financing. And I don’t know if there was the blue-red consideration there at all, but he was very unhappy when the New York MTA took out some debt at the municipal liquidity facility. There was no secondary facility for munis. Again, it’s a different market, but safety-wise, there’s no reason it should have been treated so differently.

Beckworth: I imagine maybe in his mind and maybe people around him, they were thinking, “We’re not going to bail out some state’s pension program,” but this is very different than that, right?

Kelly: Yes, I think that’s certainly a big piece of it. The Fed thought about doing something for munis even in 2008 and decided this is politically toxic or too dicey. Certainly the politics of local finance are even worse at the Treasury.

Beckworth: Okay. Now, I guess the rest of the story for this period, before we move on, is that he also proceeded to pull all the funds from the CARES out, back from the Fed, and he placed it in the Treasury general account where it couldn’t be tapped. He wanted to make sure it wouldn’t be used in the future for these facilities. Moreover, I don’t know if he encouraged, but there was legislation passed that said you couldn’t use the ESF funds for these facilities.

Kelly : This was another important part of this saga, where the timing politically was lost on nobody, too. After Trump loses his reelection bid in 2020, his Treasury Secretary Mnuchin moves quickly to close the 13(3) facilities, or he says he’s not going to renew basically all the CARES Act facilities.

He does this controversial move where he pulls the already allocated money back, citing a very wobbly interpretation of the CARES Act. He had the strength of being able to say, “Look, I was in the room when we wrote the CARES Act. I know what it means.” Basically nobody agreed with him. The Congressional Research Service said it wouldn’t hold up in court. Again, even just reporting on it was like the law does not say this. Used the controversial interpretation of the law and said we’re going to take the money that hasn’t been lent in these facilities back.

The facilities were going to “leverage” Treasury money, 10 to one, 14 to one, eight to one, whatever. The way he decided to pull money back was, all right, we’re going to leave dollar for dollar Treasury money in these facilities because they ended up not having to lend that much. They functioned as backstops. There was $13 billion in the corporate bond facility that had been purchased, and so he left $13 billion. People at the Fed are surprised.

That tells you that this was sudden and this was a new decision. It wasn’t baked into the law necessarily where it was planned for. Everybody’s freaking out at the Fed and it’s a surprise and whatever. Then there’s a new administration. They go, “Okay, well, maybe we can restart these things. We’ll put the money back in. He left $13 billion in this facility that we were going to leverage 10 to one. Maybe we can do $130 billion of lending in that facility.”

The lame duck Congress gets wind of this and basically says, “All right, there’s no way we want to let this happen.” First they go after 13(3) and they’re going to say they’re going to change 13(3) to not allow any facilities that are similar to these CARES Act facilities at any point in the future. That gets watered down. Now the law ends up saying the Treasury cannot use the ESF to invest in facilities that are the same as these new facilities that came around in 2020.

That was the outcome. It was a real scare for a while that 13(3) was going to get altered. Again, Congress just backed their man and went after these Fed facilities. It was all this drama over what was a backstop facility. What made it so dramatic is, the Fed likes to play ball, and it did with Mnuchin’s grab of the cash back, but they put out a statement to the press that says the Fed would prefer to keep these facilities open, but the secretary has indicated his reading of the law and we’re going to send the money back.

Again, this is end of 2020. No vaccine. Very uncertain future economically and from the pandemic. The Fed is worried that Mnuchin doing this is going to send the message to the markets of, oh, their support is over. Again, you see the interference of politics and this Treasury money with what the Fed wanted to do and the Fed’s messaging, which was we are still backstopping the economy.

Beckworth: Yes. That statement from the Fed was very surprising but understandable too at the same time because they wanted to communicate to the market. This just, again, illustrates the bigger point of your paper that if you’re going to go to the Treasury for equity funding, there’s a bunch of baggage that comes with it. Maybe it works out okay. Maybe it doesn’t. As you said, there’s no law that says they have to go to Treasury. If you do, the president and Treasury secretary may not be on your side. They may be pushing in another direction.

Banking Turmoil of 2023

Let’s move to the banking turmoil of March 2023. This facility the Fed introduced, now it’s a 13(3) facility. It looks and sounds a lot like a discount window facility. Maybe help us understand that.

Kelly: The Bank Term Funding Program was the 13(3) facility the Fed rolled out after the failure of SVB and Signature. It was only for banks, much like the discount window. It offered one-year loans. The discount window can only go up to four months. It offered one-year loans, and it valued Treasuries and agency collateral at par. This was a big deal because the Fed had just raised rates very quickly. The securities books of the banking system and the loan books were underwater. This overvalued Treasury collateral with respect to the loan. What it was effectively doing is lending against the value of banks’ deposit franchises. Banks don’t actually pay the market rates that are represented in mark-to-market pricing of Treasury securities.

They pay zero on our checking accounts, and they pay something less on our savings accounts, less than Fed funds rate. This really lent against the deposit franchise value, but that mechanism was to overvalue the “Treasury securities” and agency securities to provide a higher-value loan and avoid what was perceived as the Silicon Valley Bank crisis, where they effectively had negative equity when you marked their asset, their agency debt, their safe debt to market. They effectively had negative equity.

Beckworth: You just answered the question I was going to give you, and that is, some people say this was giving a transfer to the banks. But your whole point is, well, actually, they had an implicit asset behind that. That is the franchise of the depositors.

Kelly: You, the Fed, are standing behind the collective deposit franchise value of the system, and banks still have to be solvent, access a 13(3) facility. There’s a question of whether First Republic really was, but aside from that, we don’t have a record of any banks failing with a BTFP loan. Republic First had some borrowing as well. That’s not First Republic. It’s Republic First. Broadly speaking, this facility lent $165 billion, and it seems like that money is all coming back.

Beckworth: That answers that question. But how does this experience, again, speak to this dependency or this desire to quickly run to Treasury? Give me some equity.

Kelly: Again, the Bank Term Funding Program, so the collateral is worth less than the loan. That’s true, but what the Fed has is a senior claim to these solvent banks’ balance sheets. In addition to the Treasury collateral itself that it got, which at least from a credit perspective is safe, the Fed has also got a super senior claim on solvent banks’ balance sheets. If a bank defaulted with a $100 BTFP loan backed by $70 of Treasuries, the Fed still has a senior claim ahead of the FDIC for that remaining $30.

Again, you, as the Fed, are staking this program on there’s value in deposit franchises. From a systemic level, we’re going to backstop that. Again, there’s value in the pool that you’ve created of we’re lending to solvent banks. There’s earnings associated with that. The program got $25 billion from the ESF. Shows up nowhere on the ESF balance sheets, by the way. It’s in the press releases, but that’s it. The program ends up doing $165 billion of lending. This $25 billion is, say, one-sixth of its ultimate lending.

It stands to reason that even with that $25 billion of equity, the collateral that was provided plus the equity does not necessarily equal 100% of what the loan value was. We’ll find out in two months when the transaction data comes out. It wasn’t as though Treasury said, “We will give you $1 of equity for every dollar that the collateral is worth less than the loan.” There wasn’t a science to it, so much as it was $25 billion, go do the program.

It’s not clear there were huge drawbacks in this case, but it seemed like almost muscle memory. “Oh, we’ll go get some Treasury money.” We’re sitting here in the first week of a Trump administration. If you told me, what are the odds that the $215 billion in the ESF is still there in four years? I wouldn’t say 100% for sure. That’s tempting money, and it’s always referred to as a slush fund by whoever the minority party is.

That money shouldn’t be relied upon either as always being there. You don’t want the Treasury to turn down funding or to invite some sort of meddling. There was reporting before when Trump was still a candidate that his advisers really wanted a more muscular role for Treasury, particularly in these Fed programs where Treasury was helping them. Adding Treasury money, again, is only going to invite that kind of meddling.

Beckworth: The Fed was fortunate in that there wasn’t a lot of politicization or strings attached, but it can’t bank on that being there in the future, and if anything, it may be less. I also think it’s ironic that we just talked about how Congress said, “No more can you use ESF funds for CARES Act facilities.” “Well, okay, we’ll do a different facility. We’ll do this version, which is different than anything else before.” When you made that facility, you had to be mindful that you didn’t do something similar to what’s done in the CARES Act.

Kelly: You just couldn’t do something that was “the same as.” That’s the language. The question is, what does “the same as” mean? Does it mean identical? Can we just change the acronym and it’s a different facility? That’ll be the challenge for a future Fed.

Advice for the Fed

Beckworth: Tell us, what would you tell the Fed then, going forward? What approach would you tell it to take to wean itself off of this tendency and to be more effective?

Kelly: I would say start with the standard in the law, which is Treasury secretary must approve. Not Treasury secretary must provide funds. That’s step one. Step two, I would say, is any time you are thinking of taking Treasury funds, make sure there’s a clear agreement with Treasury in advance that the money is not coming back. Because really absent that, you don’t need Treasury money. This Treasury-Equity Model could work in expanding the Fed’s risk capacity.

The Fed to be secure to its satisfaction, must expect full repayment. If it has $50 billion of Treasury money, it can move to the left on the loss curve and expect $50 billion of losses. Really making that expectation clear. If there’s not agreement, saying you don’t need the money. 

Now, that’s not to say all the politics go away. You still have to get Treasury secretary approval on a term sheet. When there’s no Treasury money, it seems to desensitize the Treasury to this a bit and reduce the levers to meddle with things later. The third thing would be nominal GDP level targeting. That would be the big one. That’s just table stakes.

Beckworth: Absolutely. Pure genius there at the end. With that, our time is up. Our guest today has been Steven Kelly. Steven, thank you for coming in for the show today.

Kelly: Great to be here.

Beckworth: Macro Musings is produced by the Mercatus Center at George Mason University. Dive deeper into our research at mercatus.org/monetarypolicy. You can subscribe to the show on Apple Podcasts, Spotify, or your favorite podcast app. If you like this podcast, please consider giving us a rating and leaving a review. This helps other thoughtful people like you find the show. Find me on Twitter @DavidBeckworth and follow the show @Macro_Musings.