Per Åsberg Sommar on the State of the Riksbank and Operating Systems Around the World

Per Åsberg Sommar is a senior advisor in the markets department at the Swedish central bank. In Per’s first appearance on the show, he discusses his career as a central banker, the history of the Riksbank, evolutions in inflation targeting at the Riksbank, changes in the Sweden’s central banks operating system, its new tool called the Deposit Requirement Facility, and much more.

Subscribe to David's new Substack: Macroeconomic Policy Nexus

Read the full episode transcript:

This episode was recorded on December 12th, 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 Per Åsberg Sommar. Per is a senior adviser in the markets department at the Swedish central bank, the Riksbank, and currently is working in the research division. Per joins us today to discuss the Riksbank, its operating system, and the broader push among some central banks toward demand-driven operating systems. Per, welcome to the show.

Per Åsberg Sommar: Thank you very much. I’m very glad you’re having me here.

Beckworth: It’s great to have you on, and I have to tell the listeners the back story to getting you on the program. I was having a conversation with our friend Bill Nelson, who you also know, and another friend, Stephen Kelly, and we were discussing operating systems at central banks online via email, and we started talking about the Riksbank, and particularly this new deposit facility that you have there, where banks have to put some of their deposits and earn 0% interest as a way to build up the capital of the bank, should it be needed.

It was really interesting, so we were debating is this purely just to build the equity up, or is it also to find a way to replace the currency franchise? Bill Nelson said, “Let’s just ask the person who knows this. Let’s email Per directly.” We emailed you, and we started talking to you, and we thought this would be a great conversation to have with the rest of the listeners. Thank you for engaging us at that time.

Sommar: No problem. Thank you very much.

Beckworth: No, I know you have to give a qualifier statement before we get going.

Sommar: Yes. What I’m saying here is the views of my own and not the ones of the Riksbank.

Per’s Career

Beckworth: Okay. Now, Per, tell us a little bit about yourself. Tell the listeners what you do, and tell us how you got to the Riksbank.

Sommar: I graduated in 1998 from Uppsala University, a PhD in economics, focusing on taxation, public finance, and the effects of housing prices of the Swedish tax reform. Then I ended up, after a couple of years at the university, at the Ministry of Finance for a year. Then I started at the Riksbank in 2000 and have been here since then. I started at the financial stability department, going over to the monetary policy department, and then to the markets department. I’ve been head of the division that was responsible for the operational framework. The introduction of transaction-based reference rates was there, and also the legal issues within the markets department’s areas of responsibilities. Since November this year, I’m doing research, mainly focusing on operational frameworks and short-term money markets.

Beckworth: You’re in the research division, and I’ll give a shout-out to our friend, Roberto Billi. He’s someone that I interact with online. I know he’s there at the bank, so I’ll send my greetings to him. You’re a neighbor there in the research division. 

It’s great to have you, Per, because you are someone who actually has worked on, devised, processed central bank operating systems. You provide a fresh perspective, something we don’t have here in the US or maybe at the ECB. We recently had someone from the Reserve Bank of Australia talk about their perspective. We’re excited to hear your story, the Riksbank story, about operating systems.

Riksbank

Now, let’s talk about the central bank itself for a little bit. My understanding is the Riksbank is the oldest central bank currently in operation. Tell us a little bit about the back history to your bank.

Sommar: Yes, that’s true. The Riksbank is the oldest central bank existing. It was established in 1668, but it had a predecessor, actually, and that was Stockholms Banco, that was founded in 1656 when Johan Palmstruch got granted the right to run a bank. That bank, Stockholms Banco, they actually received deposits from a general community. And Stockholms Banco issued notes where the holder of these notes actually had a claim on Stockholms Banco corresponding to the means of payment that they had deposited. You could call them stable heavy coins in the form of copper coins.

Eventually, there were rumors that these notes wouldn’t cover the assets in the banks and there was a run on the bank. In 1668, the bank was taken over by the Swedish parliament, and they founded the Riksens Ständers Bank that was the Riksbank. They changed the name later on. 

For the first 10 years, the Riksbank was actually not allowed to issue notes since the experience from Stockholms Banco. Ten years later, the Riksbank was allowed to issue notes again as it was more efficient to use them as means of payments rather than those heavy copper coins. Since then, the Riksbank has been going on.

Beckworth: That is so fascinating, Per, and it speaks to the rich history of your institution. In fact, it’s been around longer than the United States of America has been a country: 350-plus years. That’s just amazing. Lots of great history, stories. Something else that your bank has that I really love is you have historical data that goes way back to the 1600s. My friend, Josh Hendrickson, and I know others have used that data to test macroeconomic theories. For example, the long-run neutrality of money. You guys have money supply data going way back, all the notes that you’ve issued.

Really, a phenomenal institution, a gold mine for researchers who want to do both historical questions, but also where we are today, and that’s why you’re on the show. In the present, what do we know about operating systems? What can we learn from the Riksbank? 

Now, before we get there, just a few more points about your bank. My understanding is from 1931 to 1937, it also was a price level targeter. Not just an inflation rate targeter, but a price level targeter, which is where you make up for misses in the level, which is something that comes full circle in the late 1990s, 2000s, when a bunch of economists worried about the zero lower bound, worried about deflation in Japan, zero lower bound in Europe. 

In the US, we start talking about makeup policy, forward guidance, price level targeting, in my case, nominal GDP level targeting. Ultimately, FAIT, the Fed’s flexible average inflation targeting framework, was a variant of that thinking. Lo and behold, the Riksbank did it back in 1931 to 1937. That’s fascinating.

Sommar: There is an article by Claes Berg, Lars Jonung from 1999, where they write about this period. After the gold standard broke down, Sweden let the exchange rate float and adopted a price level target from 1931 to 1937. During that time, gradually, Sweden pegged exchange rate to first the sterling pound and then to the US dollar to have an anchor for the value of the domestic currency. Since then, the Riksbank and Sweden has had some kind of a fixed exchange regime until 1992, when the krona was floating again.

Beckworth: What’s neat is the Riksbank was also a part of the early inflation targeting adopters in the early 1990s. Now, the case of the Riksbank was 1994. Let me put a bug in your ear, Per, you can tell your friends at the Riksbank. Maybe at some point, you guys should start thinking about nominal GDP targeting as something you might consider in the future, because if you get a place like the Riksbank or the RBA, or the Bank of England, try it, eventually, the Fed may try it as well. In any event, let’s go back to the history of the Riksbank. Enough daydreaming for me. Tell us about the adoption of the inflation target at the Riksbank.

Inflation Targeting at Riksbank

Sommar: After letting the Swedish krona float, the Riksbank adopted formally the inflation target in 1995. I think Bank of Canada and Bank of New Zealand had already adopted inflation targeting by then. From 1995, the Riksbank targeted an inflation rate of 2%. Initially it applied an inflation interval, 2% plus, minus 1%, but that was later abolished in 2010 as the inflation target actually was and is 2%, and that is what the Riksbank wanted to achieve.

Beckworth: No longer the 1% to 3% band around the 2% target?

Sommar: No.

Beckworth: It’s interesting to see some countries that do inflation targeting; they do have this band. The RBA has, I believe, it’s 2% to 3%, but they want to land on 2.5%. I think the Bank of Israel also does 1% to 3%. You’re not alone, and it’s interesting to see the arguments for that. I think the argument for a band is you have supply shocks, so you want some flexibility. I think what a lot of people, and what you would argue, is that you can still keep that flexibility with a regular 2% inflation target. Over the medium run, just aim for 2%, and you get that. 

I want to ask a question about the Riksbank in terms of spillover effects. You’re a central bank, a small open economy. Exchange rates matter a lot. In fact, the Riksbank on their balance sheet has a lot of foreign exchange reserves. It’s part of its stock of assets. It’s treasure chest in case it needs to use them. How consequential are policy changes in the US? Does it come to Sweden, and you guys have to respond to that, or do you feel like the floating exchange rate does most of the adjustment for you?

Sommar: A small open economy is affected by policies in larger economies, so that’s unavoidable. 

Beckworth: I guess this is my question: Do you guys pay a lot of attention whenever there’s an FOMC meeting? What are they going to do with rates? Look at what the market forecasts are. I know you’re not a policymaker, but you’re a senior staff member. You’re thinking through these issues. Are you thinking, “Oh, dear. Here goes the Fed again”? Do these thoughts ever cross your mind?

Sommar: I think it’s unavoidable to follow what other central banks are doing, but I would like not to comment on what the implication of the monetary policy is.

Beckworth: All right, I don’t want to get you in trouble.

Sommar: Thanks.

Riksbank’s Operating System

Beckworth: We will not go any farther down that path. Let’s talk about the Riksbank’s operating system. When it adopted its inflation target, I understand you guys also introduced a new operating system. Tell us about the initial operating system that you set up. Again, the historical context is in 1992, the big exchange rate mechanism broke down. Famous story in the UK, the George Soros bet against the pound. There’s that famous story. You guys also were part of that story. As a result, you established an inflation target eventually and an operating system or an implementation framework. Tell us about that.

Sommar: Yes, actually in 1994 we introduced, or the Riksbank introduced, a new operational framework that was actually a corridor system. I would say that any operational framework where the central bank offers unlimited lending and deposit facilities is some variant of a corridor system. In ’94, the Riksbank introduced this system with a wide interest rate band of 200 basis points and targeted actually initially the two-week interest rate. 

After a couple of months, the Riksbank noticed that the overnight rate that the central bank unavoidably has to stabilize to form a stable anchor for interest rate formation on longer maturities actually was quite volatile. Therefore, the Riksbank introduced daily fine-tuning transactions where the banking system could place or borrow their net position toward the Riksbank at an interest rate 10 basis points above or below the policy rate.

In this corridor system, the Riksbank conducted regular market operations to balance the system close to zero, and then using fine-tuning transactions to handle daily fluctuations due to autonomous factors. When talking about operational frameworks also, one would like to consider what the building blocks actually are, and that is an operational target, the instruments, the counterparties, and also the collateral.

Depending on local circumstances or local conditions, different central banks may end up designing the operational frameworks in different ways, but they are all trying to achieve basically the same thing: to control the short-term interest rates, even if there is a target to stabilize longer-term interest rates.

Beckworth: Whether you’re in a floor system with ample reserves or a scarce reserve system, a corridor, a ceiling system, the goal is to have interest rate control around the targeted interest rates. You follow a Taylor rule, let’s say, and it tells you to raise interest rates by 25 basis points; you want to have great interest rate control. 

I know, for example, here in the US, there’s been some concern recently about overnight repo rates going above our ceiling facilities. Can we make them more robust? You have facilities at the top, just the ceiling facilities or the lending facilities. You have facilities at the very bottom, which would be your deposit facilities.

Certificate of Deposits

Beckworth: Per, the Riksbank has this really interesting tool, certificate of deposits, that it can use to drain liquidity. Now, these CDs are a week, and they’re different than reserves. Is that right?

Sommar: Yes, they are, because the Riksbank certificates are tradable securities issued by the Riksbank. That is a financial asset, but compared to deposits overnight, that is actually account-based money, reserves that banks can use for transaction needs. The certificates they run at the policy rate, while the deposits are remunerated at 10 basis points below the policy rate. Now, to the extent banks place liquidity with these Riksbank certificates, they can actually use them as collateral to borrow money from the Riksbank in the overnight facility at the cost of 10 basis points above the policy rate or sell them back to the Riksbank at the cost of 10 basis points on the nominal value.

Beckworth: This is a way to drain liquidity. One shouldn’t look at the size of the central bank’s balance sheet, but look at the amount of reserves or settlement balances to get a true sense of liquidity, because you might have these CDs on your portfolio.

You mentioned counterparties, collateral. There’s a lot of technical issues here, but just stepping back from all those, just for a minute, the initial system you set up, would you call it a scarce reserve system? You said corridor, but corridor can sometimes mean different things. Would you consider it a scarce reserve? You did say very few reserves were there.

Sommar: I would say that the operational framework that the Riksbank introduced in 1994 basically is the same operational framework that we operate today. The foundation for that is that the Riksbank provides the amount of reserves banks demand given the current interest rate and pricing of our monetary policy instruments. If the central bank is not doing any other policy measures, or in the case of Riksbank, then the Riksbank offers the amount that they demand given the price. 

For different reasons, there might be a structural deficit or surplus that the banking system has toward the Riksbank. It is this surplus or deficit that the Riksbank tries to balance on a regular basis. That is to make the system scarce. Since 2015, for instance, the Riksbank has conducted quantitative easing and purchases of government bonds, further increasing it during the COVID crisis, 2020, and that caused a huge liquidity surplus. The way the Riksbank handled the liquidity surplus during those years was to offer the banks to place the liquidity surplus with Riksbank certificates at the policy rate. To the extent banks actually place the liquidity in those certificates, the system is balanced, and then you end up with a scarce system on a daily basis.

If banks don’t demand certificates and placing less liquidity with those certificates, then you end up in a more or less ample reserve system. There are ample reserves available on a daily basis. Still, it’s the pricing, and it’s a demand-driven and incentive-based system that may be varied over time to some extent, but the basic idea holds, I would say.

Beckworth: You have a demand-driven system, is what you’re saying?

Sommar: The basic is that. If we go on then to 2015, when inflation was low, and the Riksbank conducted supplementary monetary policy measures to achieve the inflation target by, in addition to lowering the interest rate, the policy rate bought Swedish government bonds to affect longer-term interest rates, then the system became very liquid, and there was a large liquidity surplus. Basically, if the Riksbank drains the liquidity with these regular market operations by issuing certificates, then the system becomes scarce.

Beckworth: A quick question here. You make this point that you can be in a scarce reserve system even if you have a big balance sheet, say, relative to before, because really it’s where you are on that demand curve for bank reserves. If you go past the flat part to where it’s upward sloping—like 2019 in the US, the Fed still had a big balance sheet. The structural demand for reserves had grown so much that even though there’s a lot of reserves in the system, the demand was great that they accidentally fell back into arguably a scarce reserve setting, so they had to quickly inject more reserves.

Is that the point you’re making, is that the Riksbank wants to be in a scarce reserve system such that banks themselves are driving the growth of liquidity in the system?

Sommar: Yes, that’s the basic idea, but if there are needs to do other policy actions that affect the banking system’s liquidity position, then we should be able to control the short-term interest rates in that environment, too. That was actually the task that took us on when we reviewed our operational framework in 2010, because then we have had this corridor system for some years. There were reasons to consider: Is it working as intended, or is there a need to reform it and improve it in some way? 

We actually ended up in a review in 2014. We had discussions with several central banks around the world, and also discussed very much to Ulrich Bindseil on what goals and principles should govern the design of a suitable operational framework. Actually, we ended up in suggestions and a reformation that was conducted in 2019 in two steps, and finalized in 2022, actually. The idea then was that we should design the operational framework in such a way that, no matter what the environment surrounding the Riksbank is, no matter whether there is a liquidity surplus or deficit, the Riksbank should still be able to control the overnight rate.

What we ended up in was that we actually narrowed the interest rate corridor and calibrated two plus/minus 10 basis points around the policy rate, and also sharpened the collateral requirements. Meaning that the banks had to use either government bonds or claims on central banks as collateral to be able to borrow from the standing facility at an interest rate 10 basis points above the policy rate. Still, we continued with the regular weekly market operations to balance the system as close to zero as possible, making it scarce to the extent that banks actually placed the liquidity in these market operations in the form of certificates.

Beckworth: What I’m hearing is your default setting or your normal time setting is a scarce reserve system. If you need to do QE, large-scale asset purchases, you still want a framework that allows interest rate control. Even if you fall into an ample reserve system, you want a really robust corridor for the interest rates. I think that makes great sense. In my mind, I think of an ample reserve system being part of doing QE during a crisis when the economy is weak, something’s happening. You need to put your balance sheet to use.

The question then is, what do you do after that, during the recovery? That’s where I would like to go back to something scarce or even a ceiling system demand-driven. Let me ask you, why does the Riksbank prefer its default setting in normal times to be scarce reserve? What are the advantages for scarce reserve systems?

Sommar: That depends on what you want to achieve and how you want to achieve it. Do you, as the central bank, want to maintain incentives for market-based solutions for banks’ liquidity handling or not? Do you, as the central bank, want to be able to control interest rates, no matter whether the banking system is in a surplus or deficit? Then there are reasons to have a corridor system rather than a floor system or an ample reserve system. If you have a floor system, the central bank is committed, in some sense, to always maintain the banking system in a surplus. No matter whether actually the needs for banks to have liquidity for the transaction purposes.

There are other reasons than actually interest rate control and liquidity handling that motivates an ample reserve, one can say. For the moment, actually, it is the Riksbank’s possession of securities and also the financing of the foreign exchange reserves in Swedish krona that motivates this large liquidity surplus. Now, the Riksbank has been conducting quantitative tightening in a fast, rapid pace, and that is actually, according to the already-made decisions, will be ended in this year. Then the Riksbank will end up with a security portfolio to be prepared to be able to be active in the government bond market with a portfolio of around 20 billion Swedish crowns.

What we mean with this demand-driven and incentive-based operational framework is that given the pricing and offering these regular market operations where we offer the banks to place money at the Riksbank at the policy rate for one week in the certificates, or place the money as deposits overnight at an interest rate 10 basis points below the policy rate, we actually leave it to the banks to decide to what extent the system should be ample or scarce.

In the system, it’s also important that the banks may use those certificates as collateral when needed to borrow from the Riksbank in the standing facilities, or they can sell them back to the Riksbank at the cost of 10 basis points, converting those securities or store of value to reserves. That is actually account-based means of payment in the Riksbank as overnight reserves then. It’s basically the alternative cost for reserves that motivates to the extent that banks actually make the system scarce or not. It also requires that the banks are operational, ready to act in the interbank market in the ways they prefer.

If it’s unsecured, secured, or repos overnight, that’s up to the banks. They should also consider our standing facilities overnight as a viable option to get access to reserves. In that way, it would work. In this transition from huge liquidity surplus, where all banks have been used to a large amount of liquidity available and less need to interact with each other, and being comfortable with placing money with the Riksbank in a safe way at a rather good interest rate, then we are meeting a new environment where the banks have to step up and lend money to each other, and also consider our standing facilities as a realistic option to use. We could say that the Riksbank is open for transactions, like the Bank of England says that we are open for business.

Beckworth: The Riksbank is open for business. Again, where you want to end up is more of a scarce reserve system where the banks are determining the amount of liquidity in the system. It sounds like you also value some interbank activity, some price discovery between banks, some banks being counterparties to each other. There’s a price signal there. How do you rate banks? There’s something useful going on there. That’s interesting. 

Let me ask this question. Once you get to this ideal place, the scarce reserve place with a very narrow corridor, what do you see as the anchor rate? You have a policy rate that you set, but you have both a deposit rate at the bottom and your ceiling rates at the top. Which of those two do you think better anchors your target policy rate once you’re in that scarce reserve system?

Sommar: The basic idea with the narrow corridor was actually that it wouldn’t make much difference whether the overnight rate is somewhat below or above the policy rate as long as it is within this acceptable tolerance interval.

Beckworth: There’s no anchor rate per se. It’s just the narrow corridor itself that’s anchoring overnight rates.

Sommar: That was the basic idea with this narrow corridor system. Depending on what the marginal cost is for banks, do they end up on a daily basis with a surplus or deficit, the deposit or the lending facility rate is the marginal cost. According to the pool model developed by Whitesell everything is working fine in the overnight market, no matter whether it’s unsecured or a secured market, and there is no stigma or other frictions or transactions costs, then banks will hold zero reserves with the central bank overnight. They will exchange money between themselves, meaning that the overnight rate should end up in the middle of the corridor.

Now, we see some frictions, and we encourage the banks to use and consider our standing facility as an alternative and a natural ingredient in their daily liquidity management. I think we have a way to go before we end up in this frictionless world, and that the banks actually exchange money or reserves in the way that they are comfortable with. It used to be the unsecured market, and we may well end up in a more complete financial market where there are different short-term markets, repo markets overnight, secured lending, or unsecured. 

I think this is a challenge going forward, but we are taking steps in the right direction, I think. We should also be aware that it’s a quite fragmented circle of counterparties where there are some big banks and some medium and small banks. The number of counterparties has doubled, actually, since 2015, more or less.

Beckworth: The nice thing about having robust ceiling facilities or lending facilities is that you don’t have to worry as much about reserves getting locked up in the big bank or some bank. Every bank can come to the Riksbank and say, “Hey, this is how much I need.” As long as it’s considered ordinary part of business. There’s no stigma, as you said. 

Quantitative Tightening

Let me go back to quantitative tightening that you guys have been doing. I’m curious, how did you do it? In the US, we just let the securities roll off until we think we get to the place where we’re just before we hit scarce reserves. The lowest possible point for ample reserve, that’s the Fed’s approach, but they just roll things off. They don’t actually actively sell. What did you guys do?

Sommar: The Riksbank has been actively selling Swedish government bonds and letting the other holdings of securities mature. Municipal bonds, covered bonds, and T-bills, they are held until they mature. That’s a passive QT then regarding those securities. They are wounded down to the extent that they mature actually. The government bonds are actually actively sold until we have a holding of around 20 billion crowns of government bonds.

Beckworth: Were any of these government bonds long-term and therefore subject to losses when you sold them?

Sommar: It’s typically shorter duration and maturity in the Swedish government bonds compared to many other central banks’ holdings. Still, the Riksbank did some losses due to this mark-to-market valuation of the holdings. The Riksbank actually made a huge loss and eroded its capital. There was a need for a capital addition that the Riksbank actually has received.

Beckworth: Let’s talk about your new tool in just a minute, because you guys had a nice innovation to help build up that equity again. I want to come back to that. Just on the losses. You actually did book mark-to-market losses. Did you also have net interest income loss, operating losses, because you had duration mismatch on your balance sheet?

Sommar: I don’t know, actually, the contribution of the assets returned, but the foreign exchange reserves is contributing with profits.

Beckworth: Overall, operating profits, but it was the mark-to-market losses that you felt?

Sommar: Yes.

Beckworth: Okay, because here in the US, of course, we have very little foreign exchange on the Fed’s balance sheet. The big thing for the Fed was the interest income was far less than the interest expenses. We’re paying out short-term liabilities, reserves, and we’re earning far less on our long-term bonds. It’s interesting that you guys, you said, “We’re just going to do it. We’re going to suck it up, buttercup. We’re going to sell off these securities, even if there’s a mark-to-market loss.” Then you had to be recapitalized. Let’s talk about this.

The Deposit Requirement Facility

Before we get to this new tool you have, I also want to bring in this other complicating factor that you have, which is so fascinating. Maybe for you, it’s not. Maybe it’s more of a problem for you, but for me, on the outside, it’s so fascinating. That is, in Sweden, the amount of currency has fallen dramatically. Currency is like a cheap way of funding the central bank. You call it the currency franchise, right? You fund at 0%. There’s some operating costs with it, but it’s super cheap, and you guys have lost that. Then you had some losses on mark-to-market. You were in a place where you really needed to find a way to increase your capital, and you wanted to do that without going to the government because then you were worried, “Oh, then we’ll lose our independence.” Tell us about this new tool, the history, how it works, all the good stuff.

Sommar: As you mentioned, the demand for notes and coins is quite low in Sweden, and households and companies they are not doing their payments using notes and coins, and they use other means of payments instead. The seigniorage is low, and this traditional contribution to earnings is quite moderate for the Riksbank, so we can’t rely on that, even if it contributes with its return on our assets. According to the new Riksbank Act that was adopted in January 2023, the Riksbank shall have a capital of 60 billion Swedish crowns to ensure its own financing and ensure its financing independence. This 60 billion should be kept up with the CPI, the inflation rate.

Currently, the Riksbank’s targeted capital should be 63 billion crowns, while the actual capital that the Riksbank has after capital addition is 23 billion crowns around. This new Riksbank Act was recently amended this year, in January 2025. That gave the Riksbank the right to require banks and credit institutions in Sweden, and also branches of foreign banks active in Sweden, to hold deposits with the Riksbank at zero interest rate to improve the Riksbank’s earnings. This is completely outside the operational framework and serves as a way to improve the Riksbank’s earning and reduce the costs for the Riksbank, and also to build up its capital.

The Riksbank is allowed to require that the bank’s credit institutions, and those branches of foreign banks to hold an amount that corresponds to the difference between the actual capital and the targeted capital, that is, around 40 billion Swedish crowns. That is what the banks now have paid into the Riksbank, and holds this amount at zero interest rate. Depending on how much short-term liabilities they have, deposits and short-term issued securities, these 40 billions are divided between those banks. That is what they hold until further notice, until new decisions are taken to improve, actually, the earnings on the Riksbank, and contributing to building up its capital.

Beckworth: It’s only on some of the deposits banks have at the bank, right? It’s just a certain portion?

Sommar: Yes. The base is the financing needs of the Riksbank. If the difference between the Riksbank’s actual capital and the required capital is 20 billion crowns, then it’s 20 billion that it’s required to hold.

Beckworth: Banks are earning some return at the central bank. Just some of their deposits have to be kept at 0% until the equity is earned back up. Now, you mentioned this is outside of the monetary policy operating framework, the implementation system. It sure seems supportive or consistent with it, because if you want to go to a scarce reserve system, this is one way to incentivize banks to demand less reserves at the central bank. Does it not at least complement the trajectory that you want the central bank to go toward on this operating system?

Sommar: I think one can isolate it from the operational framework. Still, on the margin, I believe it would increase the incentives to exchange money, reserves between the banks. One should also keep in mind that it’s around 150 credit institutions, banks that hold these interest rate-free deposits with the Riksbank, while there are 30 monetary policy counterparties that actually can take part of our monetary policy instruments, like interest-bearing deposits overnight and our market operations, currently the issuance of Riksbank certificates at the policy rate, and also the standing lending facility overnight. 

This 40 billion reduces the available amount of liquidity in the banking system. The liquidity surplus is reduced with the corresponding amount that the deposit requirement results in, but it is an amount that is kept static and unchanged over time. I would say it’s not really part of it. It’s a real financing instrument.

Beckworth: Again, your motivation for doing this, or I should say the Riksbank’s motivation for doing this, is so that it preserves its independence from the government. You don’t have to turn to the government to get recapitalized every time you experience losses on the bank. Now, I bring this up because it’s super fascinating to see how each central bank does it differently. As you know, in the United States, we have this magic account called deferred asset where the Fed just doesn’t send any remittances back to the federal government.

At the end of the day, from a consolidated government perspective, it’s the same thing. The government’s getting less revenue. Ultimately, taxpayers face a bigger deficit, and taxpayers have a bigger tax bill. One way or the other, it’s a loss somebody has to bear. 

I think your new tool, the deposit requirement facility. I bring this up for another reason, and that is the issue of stablecoins in the US. We’re just beginning this journey. It’s a long journey, possibly, but there is some concern that if stablecoins become widely adopted outside the US, where a lot of our currency is, then it will displace physical cash. The Federal Reserve would lose its currency franchise too, or at least some meaningful part of it, and it would be vastly reduced. It would lose that cheap funding that supports its balance sheet. It too may be in a place where it doesn’t have consistent funding from the currency franchise. We may be in a place where we also need to look at something similar to what you have done at the Riksbank. 

I think something that could actually fulfill this role is “skinny master accounts” that Governor Chris Waller talked about. Now, those have been introduced as a way to provide access to the Fed’s balance sheet for special depository institutions that are custodians for stablecoins. It’s very limited use, so it’s not a typical access to the Fed’s balance sheet. The talk is that these skinny master accounts would earn 0%. In spirit, very similar to your deposit facility, and I think it’d be one way to get this cheap currency funding again. Does that seem consistent with what’s happening there in the Riksbank?

Sommar: Well, you can view this deposit requirement as a substitute for lost seigniorage. It’s a way of creating a seigniorage in another way.

Beckworth: Again, that wasn’t the intention, but I think that could work out well to the extent stablecoins actually do displace physical dollars. We have yet to see whether this happens. 

Other Central Banks and Demand-Driven Systems

Let’s switch gears from the Riksbank to this broader movement around the world, at least for advanced central banks. We see a lot of advanced central banks moving toward this demand-driven operating system. In the limit, I would call it a demand-driven ceiling system. The Reserve Bank of Australia, the RBA, they would, I think, feel comfortable saying they’re going toward that.

I think the full name is demand-driven, a full allotment system. Effectively, at the RBA they want—and they’ve even said this—they want their ceiling facilities to be the anchor rate. The RBA still reports its deposit rate at the bottom of that corridor, but when it does monetary policy announcements, it actually announces the target policy rate, the cash rate, and then the ceiling rate. The ceiling rate is supposed to be the anchor now for the policy rate, even though the deposit rate’s still there, as they go toward a demand-driven ceiling system.

It’s really fascinating to see them. I think they’re the extreme case, but the Bank of England, the ECB, the Riksbank. What is your sense of what’s happening here? Why do we see this movement? We have multiple central banks on this journey. Of course, you guys are all on different stages of the journey. You’re all a little bit different, but many of them will use this term demand-driven system. What is your sense of the bigger picture? What’s happening here?

Sommar: Well, I think it’s a way of reducing the risks on the central bank’s balance sheets because there was a new ingredient when central banks bought securities, at least for the Riksbank, because up to then, the Riksbank had conducted monetary policy by offering lending and depositing with a positive net interest. By buying assets, the central banks keep assets on its balance sheet for a longer time. It also incurs some risks to the central bank. For instance, reducing its holdings of securities by QT, it reduces the size of the balance sheet and also the assets and financial risks associated with that on the central bank balance sheet.

It’s a way of not holding more financial assets on its balance sheet than needed to fulfill its policy objectives and purposes. What happens is that the assets has to be replaced with other assets, and that is through lending, I would guess. If you want to keep an ample reserve system, you need the banking system to demand the amount of reserves required to maintain an ample reserve system, if you care about the amount available. 

What is happening, I think, is that ECB and Bank of England are offering weekly market operations where they offer full allotment lending. To the extent that banks actually demand reserves, then the ample reserves regime will be maintained, but it could also end up as a scarce system if banks don’t demand the amount that an ample reserve requires. Then you could even end up in a system where there is a deficit and close the balance. That is also why I think some central banks consider to complement those lending facilities or market operations offering short-term lending around one week, three months, one month with asset holdings. That’s a way of ensuring that the banking system is in a liquidity surplus or ample reserve regime system.

Still, even if you maintain an ample reserve system, you can’t guarantee that reserves are redistributed among the banks in an efficient way because there might be banks hoarding liquidity for some reasons and not lending to other banks that has a deficit. Those banks have to borrow from the central bank. Then it’s very important that they actually view the central bank’s standing overnight facilities as well as their regular market operations as alternatives to get access to reserves. I don’t know if a central bank offers a regular one-week lending market operation, for instance, if that means that they end up in a ceiling system.

Because at the end of the day, it depends on where the reserves end up. If they end up as deposits, then the marginal cost for money is the deposit rate. I would say it depends on if the banking system is in a deficit or in a surplus at the end of the day, which interest rate is the alternative cost, and setting the interest rate in the overnight market. That might very well be the deposit rate if there is a surplus at the end of the day.

I think applying or operating a corridor system with a narrow band, given that there are strong incentives enough to exchange reserves between the banks and the standing facilities are not associated with stigma, that the banks are actually considering using the standing facilities, then you end up in a system where the overnight rate is somewhere between the overnight lending rate and the overnight deposit rate.

Beckworth: This is so fascinating. What you’ve described, if I understand you correctly, is that if you have a narrow band in your corridor that banks themselves can determine whether they’re going to be in a scarce reserve or an abundant reserve. It’s not inevitable that you end up in a scarce reserve system. Again, given a narrow spread between the top and the bottom, banks could decide to end up at the deposit facility or up lending from the top. 

Let me ask this final question as we wrap up our time. This has been so much fun, Per, but let me ask this question because here in the United States, we have challenges with our ceiling facilities. Our discount window has stigma. The standing repo facility hasn’t been used like policymakers had hoped it would. I think part of the challenge in the United States is that a lot of financial intermediation is not done through banks. Banks are important, but nonbank financial intermediaries, private credit, they’re also important avenues through which liquidity goes through the financial system here, so it’s a little more complicated. I think it’s trickier. I think proposals for central clearing at the standing repo facility would open it up to more counterparties, which may help. Do you face that challenge at the Riksbank, or are banks just willing to go to your lending facilities, to your ceiling facilities, without any worry about stigma?

Sommar: Our governor, Erik Thedéen, brought this up in his speech in September this year. What we see is that liquidity surplus is decreasing in a rather rapid pace. That means that all banks have to step up and be able to participate in the bank market in the ways that they are comfortable with. As I said earlier, if it’s unsecured, secured, and repos overnight, that’s fine. They also have to consider our standing facilities as a real option to use when needed, and treat this as a natural ingredient in their daily liquidity management.

What we have seen during the last 10 years is the number of banks being monetary policy counterparties to the Riksbank, has increased from around 15 to 30. That means that a lot of banks are used to a huge amount of liquidity available with minimum need to exchange reserves between themselves, so this is a new environment. That is a challenge and requires banks to actually change their mindset to some extent, and being prepared to borrow from each other and also being prepared to borrow from the Riksbank.

This means also that the Riksbank has to offer functional facilities that is easy and fast to use, where they can get access to reserves in a fast and efficient way, but the banks must also be operationally ready and have the capacity to pledge collateral and use our facilities. That is actually something that we communicate that they should do test transactions with the Riksbank to ensure that they have the operational capacity to borrow from us when needed.

This is very important, and this is actually something that was an important ingredient in the reformation of our operational framework: the interconnectedness between the payment system and the operational framework, meaning that intraday credit that are used during the day, they are automatically converted to overnight borrowing if the bank is not covering it with some other sort of borrowing from other bank or so.

We also offer availability of reserves, those instant payments in central bank money 24/7 all days around the year. The banks that participate in this instant payment system they are also able to access reserves whenever the instant payment system is open, and that is basically 24/7 all days around the year.

The reformation was actually to have our systems ready to, in an automatic way, handle collateral and also calculate the pricing of the use of our monetary policy instrument at a certain time, six o’clock in the evening, when the real-time gross settlement system closes, and that one opens seven o’clock the next morning. While the instant payment system is there, people can make payments all the time, and there is a short break, a short “monetary policy night” of one hour when is not possible to do any liquidity transfers during that hour.

Once the instant payment system is opened again one hour later, it is possible to make liquidity transfers from the real gross payment system to the instant payment system when needed. This is an automatic and efficient way of guaranteeing access to central bank’s reserves when needed. It’s the real-time gross settlement system that is closed at 6:00 and open 7:00 in the morning, while the instant payment system is closed for one hour.

Beckworth: With that, our time is up. Our guest today has been Per Åsberg Sommar. Thank you so much, Per, for coming on the program.

Sommar: Thank you very much for allowing me to come on this program. It was really enjoyable to me.

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.