Bank and Hedge Fund Bond Purchases Have Dampened the Effects of Riksbank's Quantitative Tightening

Economic Commentary Why have the Riksbank's quantitative tightening measures had such limited effects? Erik Andersson and Peter Kaplan No. 3 2026, May 28

Why have the Riksbank's quantitative tightening measures had such limited effects?

During 2022, the Riksbank shifted from quantitative easing (QE)—which provided economic stimulus during the pandemic—to quantitative tightening (QT). Securities were allowed to mature, and eventually sales began, shrinking the Riksbank's balance sheet. QT was simultaneously expected to contribute to a more contractionary monetary policy, primarily through rising risk premiums and a change in banks' funding mix, as deposits—a relatively cheap source of funding—were expected to decrease.

Now that QT is almost complete, it has turned out that the effects have been limited. Deposits have remained stable, and risk premiums have not increased significantly. We argue that this is because banks and loan-financed actors such as hedge funds have absorbed the supply of securities.

1 The increased demand from these actors has enabled banks' funding structure to remain largely unchanged—with a continued high share of relatively cheap deposits. Loan-financed purchases also mean that investors have had limited need to rebalance their portfolios.

If access to or the cost of financing develops unfavorably for hedge funds, their ability to hold large bond volumes deteriorates. In such a situation, this would likely manifest as higher risk premiums to attract other investors and probably also reduced deposits in the banking system.

Authors: Erik Andersson and Peter Kaplan, working at the Department for Financial Stability and the Department for Monetary Policy at the Riksbank2

The Riksbank's Securities Holdings: From QE to QT

Between 2015 and 2022, the Riksbank purchased large amounts of bonds as a complementary monetary policy measure, so-called quantitative easing (eng. quantitative easing, QE). The purpose was to provide additional monetary policy stimulus to increase inflation and bring it closer to the target, and to ensure credit supply in the economy, not least during the pandemic.3

In 2022, the Riksbank then moved to quantitative tightening (eng. quantitative tightening, QT) by first letting them mature without buying new ones and then starting to sell bonds. This was expected to contribute to investors demanding higher returns (higher risk premiums) to buy the increased supply of bonds, dampened credit lending, and also less deposits at banks.

4 So far, the contractionary effects have been limited. The interest rate differential against swaps on, for example, covered bonds has been stable or even fallen, and banks' deposits have not decreased to the extent one might have expected.

In this commentary, we argue that the limited effects of the tightening measures so far are due in part to buyers primarily being banks or investors who financed their purchases with bank loans. This implies a shift of assets on banks' balance sheets from central bank liquidity to bond holdings and lending (reverse repos). The effects would likely have been greater if the bonds had instead been purchased by non-banks and financed with existing funds in bank accounts. There are also other factors that have contributed to reducing the transmission of the tightening measures. Among other things, it may be due to the tightening measures being predictable for the market, but also that the Riksbank lowered the policy rate during the period.

In this analysis, we focus primarily on the development for covered bonds, although the Riksbank has also been active in the government bond market. It is in the market for covered bonds that loan-financed actors have been most active and thus contributed to credit growth in the banking system, parallel to the Riksbank's balance sheet changing over the period.

Quantitative Tightening Affects Banks' Deposits

Quantitative tightening works through several different channels. An important channel goes through banks' deposits.5 We focus on this in this analysis.

In connection with tightening, banks' deposits can decrease because buyers who are not banks pay for the securities the Riksbank sells with funds from bank accounts. This reduces banks' deposits.6 Banks may then need to adjust by making changes to the size or structure of the balance sheet. When deposits decrease, banks may need to replace the shortfall with some other market financing, which is often more expensive. But they can also, for example, reduce their liquid assets or

3 A comprehensive review of the possible effects of quantitative easing (QE) is given in Riksbanken (2022) Riksbank Study The Riksbank's Purchase of Securities 2015-2022. 4 By deposits, we mean money in households' and companies' bank accounts. It is a significant funding source for banks. 5 See, for example, FSR 2024:1 DEEP DIVE – How does quantitative tightening affect banks? or Kaplan and Njie (2024) 6 When the Riksbank sells bonds to buyers who are not banks, the payment is made via the buyer's bank. The buyer's bank debits the buyer's deposit account and settles the payment by reducing the bank's claim on the Riksbank. The buyer thus exchanges bank deposits for bonds, and the bank's liability side decreases in the form of lower deposits. For a conceptual review, see, for example, Armelius et al. (2020).

The Riksbank's Securities Holdings: From QE to QT

down on credit lending. These adjustments risk dampening economic activity.

Tightening measures can also give rise to portfolio reallocations among the actors who buy the bonds and thus entail spillover effects to other asset classes. If bond buyers largely finance their purchases by selling other securities, the prices of these assets are pushed down, i.e., related interest rates and risk premiums rise. Such portfolio reallocations can strengthen the contractionary effects we have described above.

So far, the tightening measures have had limited effects

During the tightening measures, banks' deposits have been surprisingly stable, see diagram 1. It increased significantly during the easing measures, when sellers of bonds were paid and the funds were deposited in bank accounts.7 But no corresponding decline in deposits has been observed during the tightening measures.

Our interpretation is that transactions have occurred parallel to the Riksbank selling and letting its securities mature, which have counteracted its effect on deposits. Among other things, loan-financed actors, such as hedge funds, and banks have increased their bond holdings to a significant extent.

8 The expected return has been good for hedge funds, and access to and the cost of loan financing have been favorable. This has made it attractive for them to build up holdings in covered bonds. At the same time, banks have wanted to replace reduced central bank liquidity with other liquid assets to maintain the liquidity coverage ratio (LCR). They have done this, among other things,

7 The correlation between deposits and the Riksbank's bond holdings during the period is not perfect as there are other factors affecting deposits, including how other types of lending develop and how the banking system finances itself. But since the Riksbank increased and decreased its balance sheet in a short time from 2020, one should be able to see a clear effect on deposits. For several arguments regarding this, see, for example, Andersson and Kaplan (2024) What drove the strong fluctuations in deposits between 2020 and 2023?.

8 In, among other things, FSR 2025:1 FACT – Concentrated and loan-financed foreign holdings in banks' covered bonds, it has been noted that loan-financed holdings of covered bonds have increased significantly since the Riksbank began tightening measures.

The Riksbank's Securities Holdings: From QE to QT

by buying bonds themselves.

If banks buy as many bonds as the Riksbank sells or lets mature, it does not affect deposits in banks. This is because the adjustment occurs by banks' assets being redistributed, where central bank liquidity from the Riksbank is replaced by newly acquired bonds. At the same time, the banking sector's liability side remains unchanged.

When loan-financed actors, such as hedge funds, buy the bonds, one can expect the same effect on deposits. This can be simply described in two steps that occur simultaneously. First, the Riksbank (and the banking system) reduces its balance sheet as the Riksbank's bond holdings fall. At the same time, banks issue a new loan to a hedge fund that uses the loan proceeds to buy bonds. The net effect

9 This follows from the fact that banks create money (deposits) in connection with lending. Unlike if a pension fund buys bonds with already existing funds—which reduces deposits—loan-financed purchases mean that new deposits are created, see, for example, McLeay et al. (2014).

10 In Diagram 6 in the appendix, one can see that it is primarily the banking system's reverse repos with foreign counterparties that have driven the total reverse repo volume from 2020 and onwards. Microdata shows that the majority of these loans are to hedge funds and other actors investing in the Swedish bond market, primarily covered bonds. For further arguments, see also FSR 2025:1 FACT – Concentrated and loan-financed foreign holdings in banks' covered bonds.

The Riksbank's Securities Holdings: From QE to QT

is that the size of the banking sector's balance sheet remains unchanged and thus also deposits.9

This interpretation is supported by data when we look at the development in the banking system's bond holdings and lending via reverse repos in the next section.

Diagram 1. Banks' Deposits and the Riksbank's Bond Holdings During QE and QT

Billions of kronor

Note. The series are indexed to 0 in January 2020. Deposits refer to deposits from non-MFIs in Swedish kronor.

Source: SCB's financial market statistics, Riksbank's weekly report

Banks and Loan-Financed Actors Have Largely Absorbed the Supply from the Riksbank's Holdings

To investigate whether banks' and hedge funds' purchases have indeed absorbed the Riksbank's bond sales, we analyze the banking system's total bond holdings and lending through reverse repos. Lending via reverse repos shows the development of hedge funds' and other financial intermediaries' bond holdings that have been financed through repo loans from banks.10

To the left in diagram 2, one sees that the Riksbank's bond holdings increased significantly more between 2020 and 2022 than the sum of banks' bond holdings and lending via reverse repos decreased. In March 2022, the difference amounted to approximately 300 billion kronor. This difference shows what contribution the quantitative easing measures have had on deposits in the banking system. The increased deposits should have generated portfolio effects in the economy where bond sellers received deposits that they then likely invested in other assets.11 An indirect effect is that the banking system reduced its long-term financing in covered bonds when liquidity and deposits increased, and instead increased its short-term financing through the issuance of certificates.12

To the right in diagram 2, it is clear that the size of the change in banks' securities holdings and reverse repos from 2022 largely corresponds to the decrease in the Riksbank's bond holdings. This co-variation suggests that banks and loan-financed actors have absorbed most of the bonds added to the market through the Riksbank's tightening measures. Among other things, banks have bought liquid assets such as government bonds, municipal bonds, and covered bonds to compensate for reduced central bank liquidity.

13 Hedge funds and other repo-financed actors have primarily bought covered bonds, but also municipal bonds and to a more limited extent government bonds.14

From when the Riksbank began quantitative tightening in early 2022 to the turn of the year 2025, foreign actors, which largely are hedge funds and other intermediaries, had bought approximately the same volume that the Riksbank had reduced its holdings in covered bonds.15 For government bonds, banks had bought more than 40 percent of the Riksbank's volumes. More than 20 percent of government bonds have been bought by foreign actors, who are likely a mix of loan-financed actors and investors who financed themselves via deposits. For municipal bonds, banks and foreign actors have increased their holdings significantly more than the Riksbank has reduced its holdings.

The fact that banks and hedge funds have played a prominent role during the tightening measures also means that spillover effects to other asset classes should have been limited. Their portfolio reallocations should have been small, as the purchases only to a small extent seem to have been financed through sales of other securities. In this way, they have partly been able to avoid broader price pressure and rising interest rates on other assets. In this way, the contractionary effects of the tightening measures have likely been further dampened.

Since deposits have not decreased during this period, banks have had

11 Andersson et al. (2022) indicate that the Riksbank's securities purchases contributed to more expansive financial conditions, which means that portfolio reallocations likely occurred during the period.

12 In the article in FSR 2024:1 DEEP DIVE – How does quantitative tightening affect banks? it is shown how Sweden's three large banks have adjusted their liability side in response to increased deposits due to the Riksbank's securities purchases. With increased deposits, they have been able to reduce the volume of long-term bonds and increase the amount of short-term certificates, while the liquidity coverage ratio, LCR, has remained relatively unchanged.

13 During quantitative tightening, banks' liquid assets, HQLA, which they need to meet liquidity rules, decrease. One way for a bank to adjust to reduced central bank liquidity is to buy other liquid assets, which is shown in Diagram 7 in the appendix.

14 In FSR 2025:1 FACT – Concentrated and loan-financed foreign holdings in banks' covered bonds, it is indicated that banks' reverse repo volumes with foreign actors primarily have covered bonds as underlying collateral.

15 The development of banks', foreign actors', and the Riksbank's holdings of covered bonds, government bonds, and municipal bonds during quantitative tightening is shown in diagrams 8, 9, and 10 in the appendix.

The Riksbank's Securities Holdings: From QE to QT

limited incentives to adjust their liability side with other longer market financing, for example, covered bonds.

16

Diagram 2. Banks' Bond Holdings and Reverse Repos During QE and QT

Billions in Swedish kronor

Note. The left series is indexed to 0 in February 2020. The right graph is indexed to 0 in April 2022. Banks' bond holdings and reverse repos have been multiplied by -1 to illustrate the difference between these series in relation to the Riksbank's bond holdings. The difference between the blue and red series thus becomes a proxy for the deposit effect from QE and QT.

Source: Riksbanken and Riksbank's weekly report

The development in risk premiums on covered bonds provides further support for this interpretation. If banks had needed to increase their market financing, risk premiums on, for example, covered bonds should have been pushed up. But as is clear from diagram 3, they have been relatively stable and even fallen, despite the increased bond supply from the Riksbank's quantitative tightening from 2022.

17

To discuss whether the tightening measures can have greater impact in the future, we now look at the development in the repo market and banks' supply of repo financing.

16 Covered bonds as a share of total financing for Sweden's three large banks have decreased between the turn of the year 2019 and 2025 from 24 percent to 18 percent. The share of covered bonds and other long market financing decreased from 34 percent to 26 percent during the same period.

17 From later in 2021, risk premiums began to rise, likely due to higher inflation and interest rates, a globally different pricing of credit risk, and changed expectations about quantitative tightening.

The Riksbank's Securities Holdings: From QE to QT

Diagram 3. Risk Premium on Covered Bond with 5-Year Maturity

Interest rate differential against swap rate, basis points

Note. The risk premium refers to the difference between the nominal interest rate and an interest rate swap with the corresponding maturity. The series is adjusted with a 5-day rolling average. The marking under 2020 and 2022 refers to the announcement of expanded asset purchases during the corona pandemic and the announcement of QT.

Source: Riksbanken

The Tightening Measures Can Have Greater Effects in the Future

Loan-financed actors such as hedge funds have thus bought a large part of the covered bonds that the Riksbank sold or let mature. This suggests that access to financing via repos has been good in recent years. But if the situation changes and the banking system restricts repo volume, new bond purchases may need to be carried out by investors who finance themselves with existing deposits. This would reduce deposits in the banking system, which increases banks' need for other market financing, which is often more expensive and can have effects on financial conditions.18 How large the effects will be is difficult to predict and depends, among other things, on how credit demand develops.

There are also signs that access to repos has worsened recently. Between 2021 and the first half of 2024, access appears to have been good, as most transactions—regardless of counterparty—were carried out at interest rates close to the policy rate, see diagram 4. But during the second half of 2024, a difference between Swedish and foreign counterparties began to emerge. Foreign actors, who account for the majority of the repo volume in covered bonds, faced higher interest rates than Swedish actors. During 2025, this difference has become significant. This may indicate that banks' supply of repos is limited.

18 This follows from the fact that the banking system has adjusted by shortening the maturity of its liabilities due to the supply of central bank liquidity and deposits. When central bank liquidity and deposits decrease, the banking system needs to adjust again, but this time by lengthening the maturity of its liabilities. They do this, among other things, by issuing covered bonds. For further arguments, see, for example, FSR 2024:1 DEEP DIVE – How does quantitative tightening affect banks?

The Riksbank's Securities Holdings: From QE to QT

At the same time, the liquidity surplus in the banking system decreased in connection with the Riksbank's tightening measures, which has likely contributed to higher repo interest rates generally because demand for liquidity increased on the interbank market.

Diagram 4. Volume-Weighted Interest Rates for Repos in Covered Bonds in SEK

Interest rate differential against policy rate, basis points

Note. The diagram refers to interest rates on repos between monetary policy counterparties and other counterparties. Swedish counterparties also include Danske Bank and Nordea. The interest rates refer to repos with a one-week maturity. This means that the series can be considered as proxies for interest rates in repo transactions as, among other things, the foreign counterparties' repos often have longer maturities.

Source: Riksbanken

Even though repo interest rates have risen, hedge funds and other financial intermediaries that finance their bond holdings via repos have so far kept their holdings relatively unchanged (see diagram 5 in the appendix). This suggests that their business model remains profitable. But if the financing cost for repos increases further, the business model can quickly become unprofitable, which could lead to loan-financed actors selling their holdings. If they start selling, a larger volume of bonds will go onto the market. If the bonds are then bought by other investors who do not need loan financing, deposits in the banking system decrease. This in turn increases banks' need for alternative financing, for example, through issuances of long-term covered bonds. This would likely lead to risk premiums on these securities needing to increase relative to other interest-bearing assets to attract other buyers.

Concluding Reflections

Our analysis indicates that banks and loan-financed actors have so far absorbed a large part of the bonds that the Riksbank has sold or let mature through quantitative tightening. This means that the composition on the asset side of banks' balance sheets has changed, from central bank liquidity to increased bond holdings and lending (reverse repos). This has contributed to banks' deposits being relatively stable. The recent rise in repo interest rates, however, indicates that the

The Riksbank's Securities Holdings: From QE to QT

loan-financed actors have gotten worse access to financing. This may mean that the tightening measures will have greater impact on deposits in the future.

Part of the Riksbank's bond holdings still remains, and is to be sold or mature in the future. This can likely still be absorbed relatively frictionlessly, if repo interest rates continue to enable profitable positions for loan-financed actors and banks demand more bonds for their own liquidity portfolios. In that case, the contractionary effects may be small or absent. But if loan-financed actors were to get worse access to financing, non-loan-financed actors may instead need to absorb the remaining bonds, which would have greater effects on risk premiums and deposits than we have seen so far during the Riksbank's tightening measures.

References

Acharya, Viral V., Rahul S. Chauhan, Raghuram Rajan and Sascha Steffen (2023), ”Liquidity Dependence and the Waxing and Waning of Central Bank Balance Sheets”, NBER Working Paper no. 31050, National Bureau of Economic Research. Andersson, Björn, Meredith Beech Österholm and Peter Gustafsson (2022), ”The Riksbank's Purchase of Securities 2015–2022”, Riksbank Study no. 2, Sveriges riksbank. Andersson, Erik and Peter Kaplan (2024), ”What drove the strong fluctuations in deposits between 2020 and 2023?”, Economic Commentaries no. 5, Sveriges riksbank. Armelius, Hanna, Carl Andreas Claussen and David Vestin (2020), ”Money and Monetary Policy in Times of Crisis”, Economic Commentaries no. 4, Sveriges riksbank. Kaplan, Peter and Momodou Njie (2024), ”What drives variable mortgage rates?”, Staff memo, Sveriges riksbank. McLeay, Michael, Amar Radia and Ryland Thomas (2014), ”Money creation in the modern economy”, Bank of England Quarterly Bulletin, Q1, p. 14–27.

Appendix

Diagram 5. Reverse Repos

Billions of kronor

Note. The data points from quarter 2, 2025 are projected using the development in the sum of MFIs' lending to ”Other EU countries” and ”Other foreign countries” in the financial market statistics from SCB.

Source: Riksbanken and SCB

Diagram 6. Reverse Repos, Swedish and Foreign Counterparties

Billions of Swedish kronor

Source: Riksbanken

Appendix

Diagram 7. MFIs' Securities Holdings in Swedish Kronor

Mi