A growing share of government and corporate bonds is now held by more price‑sensitive investors, as most central banks continued to step back in 2025 and issuance levels remained high. Despite three years of quantitative tightening, central banks remain the largest domestic holders of government bonds. Meanwhile, foreign investors, who may be more sensitive to heightened geopolitical risk, hold an even greater share of government bonds, and continue to be the largest holders of corporate bonds. At the same time, structural shifts are reducing institutional investors’ demand for long-term bonds. In several sovereign bond markets, hedge funds are playing a growing role as liquidity providers and marginal buyers of newly issued bonds. These changes in the investor base towards more price‑sensitive and leveraged investors could make markets more vulnerable to shocks.
Global Debt Report 2026
Sustaining Debt Market Resilience Under Growing Pressure
Abstract
The composition of investors in government and corporate bonds has shifted as overall debt levels have increased. This shift has made financing conditions more responsive to changes in market sentiment. Combined with rising financing needs, this can constrain issuers’ decision making regarding issuance timing, size, and maturity structure.
More price‑sensitive investors are holding an increasing share of government bonds, as many central banks continued to step back in 2025, and issuance levels remained elevated.
After three years of quantitative tightening, central banks are still the largest domestic holders of government bonds with their holdings at their highest level since 2015. Domestic central bank holdings were 20% of the total in 2025, down from a peak of 30% in 2021, but still much higher than in 2007 (9%).
Foreign investors hold the largest shares of both government and corporate bonds. They hold 28% of government and 31% of corporate bonds.
Regulations introduced following the global financial crisis have increased the costs for banks to warehouse bonds and have resulted in some short-term trading of bonds moving outside the sector. This has coincided with the increased footprint of hedge funds. While they provide much-needed liquidity and have supported demand at auctions and syndications, overreliance can increase the risks of sudden turbulence and market malfunctioning.
Hedge funds are now among the most active market participants in certain core government bond markets, while more than half of surveyed government issuers assess that hedge funds are marginal buyers in their markets. Several issuers see hedge fund activity as having a positive impact on liquidity.
At the same time, various factors have led to structurally lower demand for long-term bonds, including the migration from Defined Benefit (DB) to Defined Contribution (DC) pension schemes in certain jurisdictions, the increased attractiveness of shorter-term bonds, and increased risk perceptions regarding fiscal trajectories. This has contributed to the steepening of yield curves observed in 2025. The switch from DB to DC may also see allocations move away from government to higher-yielding corporate bonds.
Several issuers have reacted by issuing a higher share of shorter maturities. This supports their cost and risk-minimising objectives, but it may also, in part, crystallise the recent changes in the investor base. Corporate issuers are also generally more flexible than governments and can make larger and faster adjustments to their issuance strategy in response to changes in yields, yield curve shapes, and investor preferences.
Global financial fragmentation risks have become an important concern for issuers, given the importance of foreign investors in almost all bond markets. Geopolitical tensions can have an outsized impact on demand from foreign investors. Most sovereign issuers expect geopolitical risk to continue affecting both primary market operations and secondary market liquidity in government securities in 2026.
The chapter uses a combination of survey data from OECD and partner countries, including responses from debt management offices (DMOs) to the OECD Survey on Liquidity in Secondary Government Bond Markets, in addition to public data on central government and corporate bond holdings. Annex 3.A provides methodological guidance.
Global bond markets have seen fundamental transformations in the last two decades. Two major crises, in 2008 and 2020, coupled with structurally higher fiscal pressures stemming from megatrends like ageing populations, have increased market-based borrowing to levels not previously seen in peacetime (Chapters 1 and 2, OECD (2025[1])). Yet the scale of borrowing is not the only dimension that has changed. There has been a concurrent shift in the structure of these markets. Today’s global bond markets are not just significantly bigger than before the 2008 financial crisis; they are also characterised by a materially different investor base. The entry, then withdrawal, of central banks and the secular shift from bank to non-bank investors have fundamentally changed market dynamics. Cross-border flows have also evolved, and high levels of sovereign borrowing have seen the outstanding stock of global bonds shift towards government securities.
This matters because different investor classes behave very differently with respect to their trading frequency, their propensity to sell in a downturn, the consistency and predictability of their demand, their interplay with geopolitical developments, and other factors that affect the stability of debt markets (Fang, Hardy and Lewis, 2025[2]). Consequently, the composition of the investor base influences bond pricing and volatility, which, in turn, passes through to the real economy, affecting investment and longer-term growth prospects (Coppola, 2025[3]; Philippon, 2006[4]). Increasing interlinkages between investor categories over time amplify this risk (Aquilina, Cornelli and Tarashev, 2025[5]).
From the issuer’s perspective, recent shifts in the investor base, notably the rising share of more speculative, and price‑sensitive investors, can entail greater risks for market functioning and financial stability. Such investors are associated with higher flight risk in periods of turbulence, and their positions may be liquidated abruptly following sudden price moves or a tightening in funding conditions. Coupled with their lower organic commitments to markets, they may pull back during country‑specific stress episodes, weakening demand when liquidity is most needed and increasing vulnerability to shocks. The growing presence of these investors can therefore add to the risks arising from higher sovereign debt (Epp and Gao, 2025[6]). At a minimum, issuers may have to pay higher premiums to attract investors, but they may also experience periods of extreme volatility and even a lack of confidence in their markets, as has occurred in several G7 countries over the past six years. Therefore, understanding these shifts, which role certain investors are playing, and the points of vulnerability are key to well-informed policymaking.
This chapter maps developments in government and corporate bond markets globally since 2007 and discusses their implications. The present section begins by briefly introducing the main entities that are active in these markets, and some of their typical motives and strategies. A more detailed description is provided in Annex 3.B. Bond holders are broken down into 12 overarching categories: central banks, Money Market Funds (MMFs), banks and other Monetary Financial Institutions (MFIs), non-Money Market Funds (non-MMFs), insurance companies, households and Non-Profit Institutions Serving Households (NPISHs), pension, foreign and other. The latter category includes general government, other financial corporations, and non-financial corporations. Within these categories, certain key entity types are also introduced.
Key entity types that are active in government and corporate bond markets
Central banks buy bonds issued by governments and sometimes private entities in their domestic currency as part of their monetary policy toolkit, in particular when policy rates are near zero. They also occasionally purchase bonds as part of their mandates to maintain financial stability.
Money Market Funds (MMFs) are a type of mutual fund that invest in cash or low-risk cash equivalents such as short-term debt instruments. They often invest in Treasury bills and hold them to maturity.
Banks and other MFIs are domestic entities or domestically based subsidiaries of foreign entities that provide services as intermediaries for financial monetary transactions. Commercial and investment banks are the most prominent amongst these. Their main role in bond markets is as intermediaries, buying bonds from issuers and selling them to investors. Their central function is to conduct market making. Additionally, they may hold bonds as part of their non-trading assets, holding them as part of their High-Quality Liquid Assets (HQLAs) portfolios or as longer-term investments as part of their treasury function.
Non-money market funds. This category includes investment funds that actively manage funds on behalf of clients. They are characterised by higher price‑sensitivity than, for example, pension funds and insurance companies, but a greater tendency to hold to maturity than more opportunistic investors like hedge funds. Exchange Traded Funds (ETFs) comprise a basket of investments usually made up of equities and/or bonds. They are traded on a stock exchange, and most ETFs track the performance of an index. They can include dedicated bond ETFs and have become an increasingly popular way for retail investors to get exposure to bonds. Hedge funds, which tend to buy and sell bonds to profit from changes in prices or discrepancies in prices between similar assets. Unlike pension and insurance funds, they are less likely to hold bonds for liquidity or yield. Instead, they are generally trying to generate high risk-adjusted returns, using leverage, derivatives, and short-term positioning, taking account of factors such as carry, roll, and volatility.
Insurance companies typically hold bonds to maturity to hedge their liabilities. They are generally heterogeneous in terms of duration preference, with life insurers favouring ultra-long duration, while the duration needs of other insurance companies (property, health, automobiles, home, liability) tend to be shorter. They tend to have relatively predictable claims patterns in aggregate, which makes bonds an attractive matching asset. Equally, regulatory frameworks typically incentivise insurance funds to hold lower-risk bonds by assigning minimal capital requirements; conversely, they disincentivise riskier assets by imposing capital charges that lead insurers to hold significant amount of capital, generating an opportunity cost that diminishes the asset’s yield advantage (Rousova and Giuzio, 2019[7]).
Households and Non-Profit Institutions Serving Households (NPISH) buy bonds for reasons including wealth accumulation, retirement planning, and in some jurisdictions for tax efficiency. Bond holdings serve to ensure some return on capital without too much risk exposure; retail investment in bond markets can therefore often be viewed as a substitute for savings in bank deposits, but with some additional remuneration.
Pension funds have long-term investment horizons and highly predictable payout schedules, and typically hold bonds to provide certainty about their future cashflows. Because of this, they tend to be less price‑sensitive than other investor groups, and typically hold bonds to maturity, favouring higher duration and, in certain jurisdictions, inflation-linked instruments. For the purposes of this report, they are split into Defined Benefit (DB) pension schemes, which provide a fixed, pre‑established benefit for employees at retirement; and Defined Contribution (DC) pension schemes, where the timing and volume of contributions are defined, and individuals carry both asset value risk and longevity risk. DB plans generally favour long-term bonds as a hedge for their liabilities, while DC plans generally hold shorter-duration bonds than DB plans and are generally considered to be more yield sensitive.
Foreign includes all categories covered here that are domiciled outside the jurisdiction in which the bond issuer is domiciled. This category includes foreign central banks and sovereign wealth funds that are the portfolio managers of other states’ financial assets denominated in foreign currencies. While central banks mostly hold bonds for exchange rate and reserve management, typically in reserve currencies (mostly US dollar, but also euro and yen), sovereign wealth funds buy bonds to accumulate and preserve wealth.
Other includes all domestic investors not covered elsewhere. This category can be grouped by general government, all non-financial corporations, and all other financial corporations not mentioned above.
How key entities typically interact with bond markets
A key consideration is how each entity typically interacts with bond markets. While each will have its own particular motives and strategies, it is useful to consider how different entities typically behave in terms of three key criteria: (1) propensity to hold bonds to maturity (or to sell them on the secondary market); (2) appetite for duration and where on the curve they typically trade; and (3) yield sensitivity – i.e. how much their demand for bonds is determined by prevailing yield levels.
The propensity of the investor base to hold to maturity will impact the balance between stability and liquidity in bond markets. Ideally, issuers want both, with a balanced distribution of hold-to-maturity investors buying right along the curve as a source of reliable demand, and also investors who increase trading volumes through regular buying and selling in the secondary market, which brings greater liquidity. Hold-to-maturity investors can also promote stability in markets during periods of turmoil, unlike more active investors who may be more responsive to market movements, which can amplify market strains (such as hedge funds and other clients of dealers selling bonds during periods of market turmoil). However, hold-to-maturity investors may also reduce liquidity by effectively reducing the share of outstanding securities available for trading.
The appetite for duration matters as it affects the instrument composition that governments and corporates can issue. One of the biggest reasons for issuers issuing higher shares at shorter maturities in recent years, as discussed in Chapters 1, is the changing characteristics of the investor base, including a reduced appetite for longer-term bonds.
Yield sensitivity is defined here as the degree to which a sector would increase (reduce) its bond holdings given an increase (reduction) in bond yields. A useful way to conceptualise this is through a yield change arising from a supply shock. An increase in bond supply, assuming no exogenous change in demand, will lead to a new market equilibrium characterised by both lower prices (higher yields) and, to some degree, a shift in the investor base (Box 3.1). All else being equal, investors who are more yield sensitive will buy or sell more of their holdings than less price‑sensitive investors in response to changes in yields.
The yield sensitivity of the investor base matters, as it affects the prices issuers can achieve when selling their bonds. Although yields are driven by a multitude of factors – mostly expectations for inflation and monetary policy for sovereign issuers, and default risk and illiquidity for corporates – overreliance on more price‑sensitive investors implies insufficient structural demand from less-yield sensitive investors. It also matters because these buyers typically might be regarded as less reliable, as they will only buy at certain yield levels but will stop buying or buy less when yields fall below this level. Retail investors are a good example of this, as they typically buy for yield, not liquidity. They have been buying a growing share of government bonds in the last four years as yields increased and remained elevated, having had largely negligible holdings prior to 2022, when yields were much lower.
Figure 3.1 provides an indicative overview of the characteristics of each entity against these three criteria. As investor characteristics differ across regions and time, and sometimes cannot be defined with any certainty, it is used here as a conceptual framework in a relative manner, rather than as a definitive statement.
Different entities have different “natural habitats” in terms of their propensity to hold bonds to maturity, their duration preference, and the price sensitivity of their demand
Notes: This figure represents a general illustration of the characteristics of different investor types. Actual characteristics will be determined by organisations’ individual trading strategies and wider market and demand conditions.
* Duration appetite here refers to the average duration appetite. Certain entities, such as domestic central banks and hedge funds, trade and/or hold right across the duration curve.
** Hedge funds are more sensitive to relative rather than outright yield levels; they typically trade price dislocations on the curve, around supply events, between markets, and between cash bonds and derivatives.
The following section describes the evolution of the investor base for government and corporate bonds in the Euro area, Japan, and the United States. The focus on these jurisdictions reflects transparency and data constraints rather than an analytical choice. Moreover, the underlying data sources’ classification and coverage are not fully compatible. The trends shown here should not necessarily be extrapolated to other countries.
The dataset is constructed following the methodology shown in Annex 3.A and holdings are broken down into the 12 categories discussed at the beginning of this chapter. In 2025, the dataset encompasses USD 53 trillion of government bonds and USD 25 trillion of corporate bonds. US bonds account for 55% of the total, split between 41% government and 14% corporate. The Euro area represents 34% (17% government, 17% corporate), while Japan accounts for the remaining 11% (10% government, 1% corporate).
Developments in the aggregate investor base
In aggregate, global bond markets are dominated by government issuance. On average, government bonds have accounted for 67% of the combined total outstanding in each year since 2007 (Figure 3.2, Panel A and B). As such, the changes in the investor base have been mainly characterised by changes in government bond markets, and in particular the increasing role of central banks since 2007, and the declining role of insurance companies, and banks and other MFIs (Figure 3.3, Panel C).1
The largest change in the combined investor base for government and corporate bonds since 2007 is holdings by domestic central banks (Figure 3.2, Panel A and B). These were around 6% pre‑2008 but have increased significantly over the years, reaching a peak of 22% in 2022. Central banks have been the single largest domestic investor category in every year since 2016, including 2025.
Higher central bank holdings are a direct result of quantitative easing (QE) programmes. The US Federal Reserve (Fed) conducted three rounds of large‑scale asset purchases between 2008 and 2014 to ease the monetary policy stance. The Bank of Japan introduced its quantitative and qualitative monetary easing in 2013 and enhanced it in the following years to try to reach its inflation target. The ECB launched the Public Sector Purchase Programme in 2015 in response to the downside risk to the medium-term outlook on price developments. This was followed by the Corporate Sector Purchase Programme in 2016. All three central banks boosted or reintroduced asset purchases following the onset of the COVID‑19 pandemic in 2020.
Central banks’ holdings have decreased from a peak of 22% of the total in 2022 to 15% in 2025, as many of them have undertaken quantitative tightening (QT) to help combat higher inflation and reduce the sizes of their balance sheets (Figure 3.2, Panel C). However, their share of holdings is still far higher than the pre‑2007 level, while the absolute value is around 66% higher than in 2015. If central banks slow or halt QT going forward, they will likely remain significant holders of government bonds in many countries for the foreseeable future.
Central bank holdings of bonds increased significantly between 2015 and 2022, but have begun declining since, though they remain the single largest domestic holders
Note: Values as of Q2 2025 for the Euro area and Q3 2025 for the United States and Japan. Data on Euro area government and corporate bonds are available from 2013 and 2015, respectively. Data on corporate bonds refers to both financial and non-financial corporate bonds. Households include NPISH. For the United States, households’ holdings of government bonds have been adjusted for estimated foreign hedge fund holdings as explained in Annex 3.A. US households’ corporate data includes hedge funds. Intra‑Euro area holdings are considered domestic. In Panel A, the dotted line represents the ratio between government bonds and the sum of government and corporate bonds.
Source: Bank of Japan; European Central Bank; US Federal Reserve; and OECD calculations.
Other than central banks, only non-MMFs and foreign investors held a significantly higher share of bonds in 2025 compared to 2007. Between 2007 and 2025, their shares increased from 5% to 11% and from 24% to 29%, respectively. The holdings’ share of every other group considered in this analysis has either remained stable or decreased. Foreign investors were consistently the largest holders over the period, accounting for between 25% and 30% of the total in every year.
The shares of bonds held by MMFs and households followed a similar trend between 2007 and 2025. Their holding shares dropped in the aftermath of the global financial crisis as interest rates reached very low levels, bottoming at 2% for MMFs and 3% for households. Their shares have rebounded since 2022. However, while the share of MMFs (5%) is now 1 percentage point (p.p.) higher than in 2007, that of households (6%) is 2 p.p. lower. It is important to note that this analysis does not consider households’ indirect holdings via investment funds, which are included in the MMF and non-MMF categories.2
The share of pension funds’ holdings of domestic bonds halved from around 8% in 2007 to 4% in 2025 (Figure 3.2, Panel C).3 This trend can be partially explained by the shift of capital allocation away from bonds since 2008 due to lower yields, changed regulations that allowed some pension funds to invest more in riskier, potentially higher-returning assets, and changes in default investment strategies (OECD, 2024[8]). Notwithstanding this, the role of pension funds in bond markets is significantly understated when ignoring foreign holdings. For instance, in 2024, pension providers in nine jurisdictions for which data are available allocated an average of 44% of their USD 4.8 trillion assets abroad. More generally, pension providers from 14 major jurisdictions invested 35% of their USD 50 trillion assets in bonds and bills (OECD, 2025[9]). This issue is addressed in greater detail later in this section, drawing from a survey conducted specifically for this report.
The diverging investor bases for government and corporate bonds
There are notable differences in the investor bases for government and corporate bonds throughout the period considered here (Figure 3.3). In 2025, foreign investors held the largest shares of both. They held around a quarter (28%) of government bonds but around a third (31%) of corporate bonds. Central banks and MMFs together held around one‑quarter of government bonds but almost no corporate bonds. This 25 -p.p. difference closely mirrors the difference in the shares held by insurance companies, non-MMFs, and banks and other MFIs. The shares held by these three categories are significantly higher in the corporate bond market than in the government bond market.
The investor base for corporate bonds has been more stable than that of government bonds over this period (Figure 3.3, Panel B). This is also a result of much smaller direct central bank interventions in corporate markets compared to government bond markets. Foreign investors, insurance companies, and banks and other MFIs held around two‑thirds of corporate bonds every year over the period, with foreign investors holding 31% on average. The shares held by banks and other MFIs declined in the aftermath of the 2008 financial crisis from 15% to 9% in 2014 but quickly rebounded and have remained stable at around 15% since 2017.
The investor base for corporate bonds has been more stable than government bonds since 2007
Note: Values as of Q2 2025 for Euro area and Q3 2025 for the United States and Japan. Data on Euro area government and corporate bonds are available from 2013 and 2015, respectively. Data on corporate bonds refer to both financial and non-financial corporate bonds. In Panel A, US households’ data have been adjusted for estimated foreign hedge funds’ holdings as explained in Annex 3.A. US households’ corporate data include hedge funds. Intra‑Euro area holdings are considered domestic.
Source: Bank of Japan; European Central Bank; US Federal Reserve; and OECD calculations.
The share of corporate bonds held by pension funds has remained stable at around 4% since 2015 (Figure 3.3, Panel B). However, as previously mentioned, this figure likely understates their holdings due to the share accounted for by foreign holdings. For example, 41% of US pension funds’ corporate bond portfolios were invested in foreign securities at the end of 2024 (Board of Governors of the Federal Reserve System, 2025[10]). In smaller countries, the foreign allocation can be very dominant: the Dutch ABP fund, for instance, invests only 7% of its corporate bond portfolio in Dutch bonds, with 64% of total bond holdings outside the euro zone, as of the end of June 2025 (ABP, 2025[11]; 2025[12]).
Households, MMFs and non-financial corporations held negligible shares of corporate bonds in 2025 (Figure 3.3, Panel B). Households and MMFs are the only categories whose shares have consistently declined since 2011, with the former falling from 10% to 3% in 2025 and the latter hovering around 0% since 2014. Non-financial corporations – included in the other category – held no more than 1% of the corporate bond market in the period considered (Figure 3.3, Panel B). However, financial disclosures suggest that some non-financial companies have major corporate bond holdings that national accounts classify as holdings from financial corporations because they are held through financial subsidiaries or holding companies (Çelik, Demirtaş and Isaksson, 2020[13]).
Sectoral price sensitivities (also referred to as demand elasticities) are a key factor when assessing the investor base and its interrelation with market developments. Sudden increases in debt are typically absorbed by market participants characterised by higher price sensitivities, resulting in a changed investor base composition. On the other hand, a lack of price‑sensitive investors ready to increase their holdings will lead to yields reacting more sharply to supply shocks. While sectoral sensitivities can be assessed qualitatively based on theoretical reasoning and anecdotal evidence, quantifying them based on empirical data is challenging. Only in recent years has an academic literature emerged applying the demand system asset pricing approach of Koijen and Yogo (2019[14]) to government bond markets.
Estimating demand elasticities is challenging in various aspects. First, market prices and quantities are determined simultaneously. It is therefore necessary to rely on methods that attempt to differentiate between cause and effect under plausible assumptions. Second, the actual relations may be state‑dependent, non-linear, conditional on unobserved factors, and unstable across regions. Third, while prices immediately react to news, changes in the investor base might follow more slowly. Fourth, shocks to one market spill over to other markets, thereby diluting the effect that would exist in isolation.
As shown in Table 3.1, there is considerable variation in the results of the empirical literature due to the challenges outlined above. Despite this heterogeneity, the following observations can be made. The estimated demand elasticity of banks and foreign private investors is often among the highest, whereas it is often among the lowest for pension funds, insurance companies, and foreign official institutions. The picture is mixed for mutual funds and households. Especially remarkable are the negative values found for pension funds and insurance companies in the aggregate Euro area and in the Netherlands. This might, firstly, point to the higher role of long-duration assets for these sectors in comparison to the United States and, secondly, to the phenomenon that liability-driven investors might be forced to increase their long-term bond holdings pro-cyclically when yields fall (Domanski, Shin and Sushko, 2015[15]).
(Percentage increase of the sector’s demand given a 1 p.p. increase in the 8y yield)
Note: Grey shading signifies that the estimate is not available or statistically insignificant at commonly used levels. Some estimates are scaled to eight‑year yield elasticities to make them comparable across the studies. For the United States, households include hedge funds.
Source: Chaudhary, Manav, Zhiyu Fu, & Haonan Zhou (Chaudhary, Fu and Zhou, 2024[16]): “Anatomy of the Treasury Market: Who Moves Yields?” Olin Business School Center for Finance & Accounting Research Paper No. 2024/14. Domanski, Dietrich, Hyun Song Shin, & Vladyslav Sushko (2015[15]): “The hunt for duration: not waving but drowning?” IMF Econ. Rev. 65.1. Egemen Eren & Andreas Schrimpf & Dora Xia (2023[17]): “The demand for government debt” Bank for International Settlements, WP 1 105. Jansen, Kristy AE (2025[18]): “Long-term investors, demand shifts, and yields”, The Review of Financial Studies 38.1. Jansen, Kristy AE, Wenhao Li, & Lukas Schmid (2024[19]) “Granular treasury demand with arbitrageurs”, NBER Working Paper 33 243. Koijen, Ralph S.J. & Koulischer, François & Nguyen, Benoît & Yogo, Motohiro (2021[20]): “Inspecting the mechanism of quantitative easing in the Euro area”, Journal of Financial Economics, vol. 140(1).Koijen, Ralph SJ, & Yogo, Motohiro (2019[14]): “A demand system approach to asset pricing.” Journal of Political Economy 127.4.
The rebound in the household share of government bonds since 2022 has not been seen in corporate bond markets. Sovereign issuers have increasingly focussed on retail investors in recent years through dedicated retail programmes, products, tax incentives, distribution platforms, and marketing efforts. Examples of new sovereign retail products issued in 2025 include the first Italian retail bond with an option to redeem at face value before maturity, and Lithuanian defence bonds available to retail and institutional investors. Several countries are also planning to issue new retail securities in 2026.4 Among them, Luxembourg intends to issue its first retail bond since 2008, aimed at solely financing defence expenditures.
As investors participate in the bond market for different reasons, their allocation preferences between government and corporate bonds varies accordingly and depends on the market in which they operate. Broadly speaking, investors in Japan and the United States concentrate most of their bond portfolios in government securities, reflecting the larger size of these markets, compared to their corporate counterparts. In Japan, this reflects a long-lasting pattern, whereas in the United States there has been an increasing trend since 2007 – driven in part by large, sustained government deficits. Euro area investors have a more balanced allocation across the two segments (Figure 3.4).
In both the United States and Japan, central banks direct nearly all their bond holdings towards the government market (Figure 3.4, Panels A and C). While this primarily reflects the key role government bond markets play in the monetary policy transmission mechanism, in the United States it also reflects legal restrictions that prohibit corporate asset purchases outside of emergencies (Fed, 2017[21]; 2017[22]). In contrast, the BoJ’s focus on government bonds also reflects the vast relative size of the government bond market, which on average accounted for 93% of the total bond market during the period considered. The share of government holdings of central banks in the Euro area, instead, ranged between 74% and 81% of their total holdings between 2015 and 2024 (Figure 3.4, Panel B), indicating sizeable holdings of non-government bonds over this period.
The trend in asset allocation of households differs across regions. While in the United States and the Euro area their share of government bond holdings is the highest since at least 2015, it is lower in Japan (Figure 3.4, Panel B). In the Euro area, their share of government bond holdings rose from 26% to 45% between 2015 and 2024. In Japan, their share declined by 34.6 p.p. from 2007 to 2025. Despite this, there has been an emerging upward trend from 2024 to 2025, as rising yields make these assets more attractive.
Among financial investors, holdings by insurance corporations, banks and other MFIs, and MMFs differed the most between regions. In contrast, pension fund allocations have gradually converged over time, largely driven by a significant increase in US pension funds’ allocation to government bonds (up by 15 p.p. in the past decade). At the same time, the share of government bonds held by pension funds in the Euro area is expected to decline further, reflecting the ongoing transition from DB to DC schemes (Figure 3.4).
Government bonds dominate portfolios in Japan and the United States, while in the Euro area the split is more balanced
Note: Values as of Q2 2025 for Euro area and Q3 2025 for the United States and Japan. Complete data on Euro area bonds are available from 2015 to 2025. Households’ data refer to Households and NPISHs. US households’ holdings of government bonds have been adjusted for estimated hedge funds’ holdings as explained in Annex 3.A. US households’ corporate data include hedge funds. A. Data on corporate bonds refers to both financial and non-financial corporate bonds. Intra‑Euro area holdings are considered domestic. Data on Euro area central bank’s holdings refers to all national central banks of the Eurosystem.
Source: Bank of Japan; European Central Bank; US Federal Reserve; and OECD calculations.
Regional trends in the aggregate investor base
The investor base composition and changes over time differ not only between government and corporate bond markets, but also between regions. While certain global trends can be observed to varying degrees at the regional level, such as the rise in central banks’ shares, other patterns, like the consistently high share of foreign holdings, are not evident in all regions. One drawback of this data is that the sectoral composition of foreign holdings is unknown. Box 3.2 provides a closer look into the development of the holding structure of German Government bonds that also takes into account the sectoral composition of foreign holdings.
In 2025, foreign investors held the largest share in the US and Euro area bond markets, accounting for 34% and 26% of the outstanding totals, respectively, whereas in Japan, the central bank held the largest share (39%) (Figure 3.5). The US market stands out due to its relatively high share of holdings by households, and comparatively low central bank holdings, with the Fed buying a smaller share of outstanding government bonds compared to other major central banks.
Central banks have increased their bond holdings across all regions over the entire period, with the BoJ’s rising the most, by 31 p.p. (Figure 3.5, Panel C). However, in recent years, these holdings have declined as central banks began unwinding their QE programmes. By 2025, the share of central bank holdings was 8 p.p. below its 2021 peak in the United States and 7 p.p. lower in the Euro area. Japan remains somewhat of an exception in this respect, with the BoJ’s holdings not peaking until 2023, at 42% of outstanding government bonds, and falling by just 3 p.p. since then.
The level of foreign holdings is affected by several factors, notably interest rate differentials, hedging costs, reserve currency status, the pace of QT, and geopolitical developments, all of which contribute to regional differences. While foreign investors have long been sizeable holders of bonds in the United States and Euro area, their presence in Japan remains limited, though it has increased gradually. Starting from 2022, foreign investors kept buying Japanese bills while decreasing their holdings of bonds, largely due to advantageous swap market pricing (Japan Ministry of Finance, 2025[23]). From 2022 to mid-2023, foreign holdings of bills increased by 13%, while bond holdings decreased by 17%. This trend started to reverse in mid-2024, likely because of higher policy rates and expectations for further hikes, moving yields higher. Also, because of these perceived higher returns, Q1 2025 was the highest quarter since at least 2021 for net foreign buying of long-term (>10 years) Japanese bonds. In the United States and the Euro area, foreign ownership has generally declined since 2015, but foreign investors remain the largest investor category.
Reflecting differences in market structure as well as regulation, the level of ownership by financial institutions differs substantially between regions (Figure 3.5). Insurance companies play a larger role in Japan, whilst banks and other MFIs remain more prominent in the Euro area. The importance of investment funds, especially MMFs, is a particular characteristic in the United States (Figure 3.5, Panel A). Domestic pension funds account for a small share of direct holdings in the United States and the Euro area. Despite holding 8% of the market in Japan, the share of pension funds in the region remains lower than the one in 2007.
Foreign investors are the largest holders in the United States and Euro area, but in Japan it’s the central bank
Note: Values as of Q2 2025 for Euro area and Q3 2025 for the United States and Japan. Complete data on Euro area bonds are available from 2015 to 2025. Household data refer to Households and NPISHs. US household data are computed as a residual category. US households’ holdings of government bonds have been adjusted for estimated hedge funds’ holdings as explained in Annex 3.A. US data on households’ corporate bond holdings have not been adjusted. Data on corporate bonds refers to both financial and non-financial corporate bonds. Intra‑Euro area holdings are considered domestic. Data on Euro area central bank’s holdings refers to all national central banks of the Eurosystem.
Source: Bank of Japan; European Central Bank; US Federal Reserve; and OECD calculations.
Investor base data usually lacks information on the composition of foreign holdings. As this sector consists of heterogeneous actors that play an important role, a more granular breakdown can provide valuable insights. The German DMO maintains a data set covering investors at a granular level, irrespective of their location. The data set is based on transaction-level secondary market data as well as data from bond auctions and syndications. The respective estimates, however, have certain limitations. In particular, banks’ holdings are likely overstated by construction, while the shares of other sectors are correspondingly understated (see methodological notes below). Figure 3.6, Panel A provides monthly data on the estimated holding structure of German federal government bonds.
Note: Notional value. Sectoral share as of 11/2025 in brackets. EMU = European Monetary Union, ROW = rest of the world. The estimated holdings are based on transactions reported by the financial institutions participating at German federal bond auctions. These transactions presumably represent most of the total trading in the market. There exists, however, an unknown number of transactions which are not reported. All issued bonds which are not reported as being sold to other market participants are assumed to be held by the category “Banks & Brokers”. This likely leads to their share being overstated. Additionally, the underlying data set begins in 01/2005. In the estimation it is assumed that the initial holding structure is proportional to the sectoral net purchases from 01/2005 to 01/2008. Lastly, a significant amount of the bonds is held by the debt management office itself. This share is excluded from the analysis.
Source: German Finance Agency.
The most noticeable observation is the large share held by official sector institutions, both within the European Monetary Union (EMU) and in the rest of the world (ROW). The EMU holdings are predominantly driven by the bond purchase programmes of the Eurosystem’s central banks. Holdings of the foreign official sector presumably consist primarily of central banks’ currency reserves. Both these sectors can be regarded as being less yield sensitive.1 Due to the ECB’s asset purchase programmes, the two sectors reached a combined share of about 70% of the outstanding amount of German Government bonds in mid-2022. Combined with the holdings of other buy-and-hold investors, this implies a low free float and depressed yields due to a scarcity premium (Paret and Weber, 2019[24]).2
While the hedge funds’ share of holdings has been low on average, the respective time series shows a remarkable pattern that correlates with the valuation of German Government bonds relative to other financial instruments. In particular, due to short selling, estimated aggregate hedge fund holdings turned negative in 2016-2018 and again in 2022-2023. These two periods coincided with German Government bonds being expensive according to several measures, such as bond-swap spreads, yield spreads relative to other issuers in the Euro area, and rates on repurchase operations (repo) using German Government bonds as collateral. This indicates that hedge funds act as an absorber of the scarcity-inducing demand by price‑insensitive investors, providing additional supply to the market by short selling.
The combination of QT by the ECB and the record amounts of additional supply of German Government bonds implies that the remaining sectors have absorbed an estimated EUR 430 billion of issuance since mid-2022. Figure 3.6, Panel B depicts the absolute changes of the estimated sectoral holdings from June 2022 to November 2025. While banks and mutual funds accounted for almost half of the increase, the shares of the remaining sectors also grew significantly. Though being lower in absolute values, pension funds increased their estimated holdings by about 74% and insurance companies by 52%. Hedge funds’ holdings grew from slightly negative values to about EUR 40 billion. Overall, it can be concluded that the additional supply of German Government bonds has been absorbed fairly broadly across sectors, albeit with an increasing role for more price‑sensitive investors.
Notes:
1. These less price‑sensitive official sectors are, however, not homogeneous. Eurosystem purchases are policy-driven, while currency reserve managers as well as public pension funds and sovereign wealth funds are likely price‑sensitive to some degree.
2. It should be noted that the scarcity of bonds is not solely determined by the holding structure but also by the availability of securities lending and repo. Additionally, security-specific scarcity and repo specialness can occur even when aggregate collateral is ample, due to idiosyncratic demand surges and intermediation frictions.
Source: Paret, Anne‑Charlotte and Weber, Anke (2019[24]): “German Bond Yields and Debt Supply: Is there a “Bund Premium”?”, International Monetary Fund, Working Paper 235.
Novel international data on pension funds’ debt security holdings
Notwithstanding a reduction in relative domestic ownership shares over time, pension funds remain important players in global fixed income markets, both domestically and internationally. Structural changes in national pension systems, notably the tendency to move from DB to DC models (Box 3.4), reverberate throughout global bond markets and have substantial impacts on relative demand between sovereign and corporate debt as well as between shorter and longer duration.
To complement official statistics and obtain more granular information on the composition of pension funds’ debt security holdings and their evolution over the past five years, the OECD conducted a survey targeting pension fund supervisory authorities across OECD countries. The survey includes data on total pension fund assets and more detailed information, specifically on the composition of their debt investments, than is available in national accounts.
A total of 33 countries responded to the survey. Twenty-three are from Europe, five from Latin America, three from Asia, one from North America (Canada), and one from Oceania (New Zealand).The survey highlights significant differences in the size of the pension sector across countries, with just three countries (Canada, Japan and the Netherlands) accounting for almost half of total fund assets in the sample in 2024, and the top six countries representing 60% (Figure 3.7).
Pension assets are distributed unevenly across geographies, and the size is not always linked to the size of a country’s economy or population
Note: The figure shows the weight of each country based on the value of its pension funds’ total assets (in USD) at the end of 2024.
Source: OECD Questionnaire on Pension Funds’ Debt Security Allocations (2025).
The aggregate asset value of pension funds covered by the survey responses ranged between USD 5.2 trillion and USD 6.1 trillion from the end of 2019 to March 2025, with the highest value recorded in 2021 (Figure 3.8). Notably, in 2022, total fund assets decreased by 15% relative to 2021, but this decline was followed by a gradual recovery in subsequent years. Despite these fluctuations in total assets, the share allocated to debt securities has remained very stable at around 47% over the years analysed.
Debt securities make up roughly half of pension fund assets, a share that has been constant since before 2020
Note: Only includes countries that reported data for all years.
Source: OECD Questionnaire on Pension Funds’ Debt Security Allocations (2025).
Similarly, the distribution of total debt assets by bond type has remained largely stable, with central government bonds accounting for just under 60%, corporate bonds around 40%, and other types of bonds (such as municipal bonds) making up the remaining 1‑2% since 2019 (Figure 3.9, Panel A). At the country level, the median allocation to each bond type has also remained relatively stable over the period, with government bonds ranging between 57% and 64% and corporate bonds between 33% and 39%. There was, however, a slight decrease in government bonds and a corresponding increase in corporate bonds in the first quarter of 2025 (Figure 3.9, Panel B). Notably, over the past two years, portfolio allocation strategies have become increasingly diverse across countries, as reflected by the greater variation in the distribution, particularly within the interquartile range.
There was a slight shift in the distribution of pension funds’ debt security holdings towards corporate debt in 2025
Note: Panel A shows the distribution of all respondents’ aggregated debt assets by type. Countries that reported only one type of bond holdings (either government or corporate) are excluded. Panel B shows the percentiles of the distribution across respondents (i.e. reporting authorities).
Source: OECD Questionnaire on Pension Funds’ Debt Security Allocations (2025).
There is a clear trend towards shorter maturities, particularly a shift from securities with maturities over 20 years to those between 5 and 20 years (Figure 3.10). This is partly a reflection of structural changes in pension systems as the gradual transition from DB to DC schemes reduces the need for asset-liability matching and consequent demand for very long-duration securities (Box 3.4). In addition, demographic changes, notably ageing populations, have increased the need for greater liquidity to meet near-term benefit payments as people retire.
Pension funds have skewed their holdings towards shorter-dated debt as interest rates have increased
Note: Refers to remaining maturities. Securities with maturities of less than one year only include government bonds. Countries without a reported maturity breakdown are excluded. Panel A shows the distribution of all respondents’ aggregated debt assets by portfolio maturity. Panel B illustrates the evolution of each shares in the overall total since 2020.
Source: OECD Questionnaire on Pension Funds’ Debt Security Allocations (2025).
Pension funds hold a significant portion of their debt portfolios in foreign securities. On aggregate, foreign debt assets accounted for roughly 60% of total debt security holdings between 2020 and Q1 2025. In 2024, within the total foreign debt portfolio, 62% was invested in Europe, 13% in the United States, and 2% in Asia (Figure 3.11). This distribution is strongly influenced by the sample composition, as most respondents are European and the Netherlands alone accounts for 59% of total foreign holdings in Europe. A comparison of the distribution of aggregate holdings by investment area between 2020 and 2024 shows that it has remained largely stable.
Foreign debt securities account for roughly 60% of total debt security holdings
Note: The figure shows countries’ foreign debt securities by investment area in 2024. Total foreign debt security holdings amount to USD 1.2 trillion. Only countries and regions explicitly reported by respondents are included; any residual, representing potential differences between total foreign holdings reported and the sum of country-level breakdowns, has been excluded. The “Other” category includes entries reported as “other” by respondents, groups of countries spanning multiple regions, and non-country issuers, such as supranational entities.
Source: OECD Questionnaire on Pension Funds’ Debt Security Allocations (2025).
The currency composition of debt holdings reflects the exposure to exchange rate risk. On aggregate, the share of foreign currency debt securities has remained relatively stable at around 25%, with a slight increase observed since 2023 (Figure 3.12, Panel A). Within these holdings, the US dollar dominates, and the share of USD-denominated holdings has shown an upward trend since 2019, reaching 54% by the end of the first quarter of 2025 (Figure 3.12, Panel B).
The share of foreign currency holdings denominated in US dollar securities increased substantially from 2019‑2025
Note: Countries that did not report currency breakdowns are excluded. Panel A shows the share of respondents’ aggregated debt assets that is denominated in a foreign currency. Panel B shows the share of respondents aggregated foreign currency-denominated debt assets that is denominated in USD.
Source: OECD Questionnaire on Pension Funds’ Debt Security Allocations (2025).
The aggregate credit rating composition of pension fund debt holdings has remained relatively stable over the last five years, with investment grade bonds accounting for around three‑quarters, non-investment grade bonds around 6‑7%, and unrated bonds between 15‑19% (Figure 3.13). However, a country-level analysis shows a different picture: the median share of investment grade bonds has steadily increased, while the median shares of non-investment grade and unrated bonds have declined accordingly. This suggests that pension funds in countries with smaller pension sectors have increasingly shifted towards more conservative allocations.
The median country’s credit risk exposure has decreased since the tightening cycle began in 2022
Note: The median corresponds to the distribution across respondents (i.e. reporting authorities).
Source: OECD Questionnaire on Pension Funds’ Debt Security Allocations (2025).
The following section presents the key drivers of changes in the investor base. These include the role of issuance, unconventional monetary policy, and the regulation of certain holding sectors.
The effects of bond supply and asset purchase programmes on the investor base
The gross supply of both government and corporate bonds is determined by the financing needs and issuance decisions of the respective entities. As laid out at the beginning of this chapter, more price‑sensitive investors will – all else being equal – adjust their holdings more markedly in response to a change in yields. Supply is a determining factor for yields; therefore, changes in the supply of bonds can change the holding structure due to the differing yield sensitivities of the holding sectors.
Bond issuance by central governments and corporates are currently at historically high levels and expected to continue growing (see Chapters 1 and 2). OECD central governments issued USD 17 trillion in 2025 and issuance is projected to reach USD 18 trillion in 2026. Corporates issued USD 6.8 trillion in 2025. This is happening in a context where central banks have withdrawn their long-standing support for markets through asset-purchase programmes, leaving a greater net supply of bonds to be absorbed by markets.
Central bank holdings of government and corporate bonds increased by an annual compounded growth rate of 11% between 2007 and 2021 in USD terms. Holdings then decreased by almost 20% over the next three years but remained stable in 2025, mainly because of US dollar depreciation. Government bonds continue to account for the vast majority of central bank bond holdings (Figure 3.14, Panel A).
Central banks are price‑insensitive bond holders whose buying and selling decisions are dictated by their policy objectives (Figure 3.14, Panel B). Between 2015 and 2019, central banks’ holdings increased in the Euro area and in Japan, while declining in the United States. This divergence resulted from different growth and inflation dynamics. Holdings increased sharply in all three cases in response to the COVID‑19 pandemic. In 2020 alone, Fed holdings doubled, while the ECB’s jumped by almost 50%.
QT policies further illustrate the price‑insensitive behaviour of central banks (Figure 3.14, Panel C). The Fed began reducing its holdings in 2022 to counter high inflation. It set a monthly cap on the volume of securities that it would let mature and roll off its balance sheet without reinvestment, first at USD 30 billion, then at USD 60 billion until June 2024, when the cap was lowered to USD 25 billion. In 2025, concerns over conditions in money markets contributed to the decisions to lower the cap further to USD 5 billion in March and end QT altogether in December (Fed, 2025[25]). This resulted in almost unchanged holdings during the year. The BoJ began reducing its holdings of government bonds in 2024 amid an improving macroeconomic environment by decreasing its monthly outright purchases by JPY 400 billion each quarter. In mid-2025, to balance market functioning and financial stability, the BoJ decided to slow the reduction in its purchases to JPY 200 billion each quarter from April 2026 (Bank of Japan, 2025[26]). The ECB, meanwhile, began reducing its bond holdings by decreasing its APP reinvestments in March 2023 and ending them altogether in July 2023. Then, in the second half of 2024, it started winding down the PEPP and stopped reinvesting altogether by the end of 2024.
Due to the magnitude of central banks’ balance sheets, their decisions to enter or withdraw from bond markets can have repercussions for the functioning of those markets beyond the intended monetary policy objective, including market liquidity and pricing. In the Euro area, for example, QE created a scarcity of government bonds that caused market frictions, ultimately resulting in significant price differences between functionally equivalent assets (Pelizzon, Subrahmanyam and Tomio, 2025[27]).
Central banks’ entry into a market can also impact the composition of the non-central bank investor base. The distribution of central bank counterparties (i.e. the selling investors) during QE was not proportional to the pre‑QE investor base, meaning QE disproportionately displaced certain investor types, notably foreign investors. In global government bond markets, foreign investors and banks saw their shares fall the most during the 2015-2021 phases of QE (Figure 3.3), while the corporate bond market in the Euro area saw an increase in holdings from within the Eurosystem during QE, which was offset in full by a relative decrease in foreign holdings (OECD, 2025[1]). Which investors central banks replace during QE also has consequences for the impact of their subsequent withdrawal from bond markets through QT.
Central banks have decreased their bond holdings since the 2021 peak through QT
Note: ECB data refer to bonds held via APP and PEPP. Data on corporate bonds refer to both financial and non-financial corporate bonds.
Source: Bank of Japan; European Central Bank; US Federal Reserve; 2023, and 2025 OECD Survey on Liquidity in Government Bond Secondary Markets; and OECD calculations.
There appears to be an asymmetry in the effects of QE and QT. In the short term, certain investor types are much more prone to absorb excess supply than others. This follows from differences in yield sensitivities, in turn stemming from differences in investment strategies and liability structures. Bond market investors can broadly be thought of either as long-term holders (e.g. pension funds and insurance companies) or liquidity traders (e.g. hedge funds and proprietary trading firms). The former group has a fundamentally larger risk-taking capacity, given its liability structures, but faces higher costs for adjusting its portfolios, whereas the reverse is true for the latter group. This means liquidity investors are more likely to buy new bonds coming to market to offset central banks’ withdrawal. However, their lower risk-taking capacity means they command higher risk premia, which results in an overshooting effect during a positive bond supply shock (i.e. QT), where short-term prices fall by more and volatility is higher than the equivalent reverse movements during a negative supply shock (e.g. QE) (Jiang and Sun, 2024[28]).
At the same time, on a macro level, evidence from the QT episode globally since 2022 suggests the impact has so far been significantly smaller than QE in reverse (Du, Forbes and Luzzetti, 2024[29]). Similarly, there is extensive evidence that QE in the US Government bond market had spillovers both internationally and to corporate bond markets (OECD, 2024[30]), but estimates suggest that the price impact on broader fixed income markets of QT in US Treasury markets has been minimal (Wright, 2022[31]).
The impact of regulations on demand and market structure
Regulation can also be a key driver of changes in the investor base. Certain regulations can be beneficial to issuers. For example, the zero-risk-weighting applied to sovereign bonds held by banks when computing their capital ratios provides an incentive to hold these over other asset types (Fang, Hardy and Lewis, 2025[2]). Equally, DB pension funds, which are subject to statutory minimum funding ratios to ensure that future payments can be met, are incentivised to invest in long-term debt securities, particularly government bonds (ECB, 2021[32]). However, the general direction post- global financial crisis has been to tighten regulations on “traditional” participants in government and corporate bond markets, such as banks, insurance companies, and pension funds.
Banking regulation
Following the global financial crisis, banks’ ability to engage in market-making has become increasingly constrained by stricter regulatory frameworks. Reforms introduced by the Basel Committee on Banking Supervision and the Dodd-Frank Act in the United States imposed higher capital and liquidity requirements, as well as restrictions on proprietary trading. Basel II.5, introduced in 2010, required banks to hold more capital against the risks of trading activities, including corporate bonds, increasing the cost of holding these assets (Adrian, Boyarchenko and Shachar, 2017[33]). Basel III, which represented a comprehensive overhaul of the prudential framework, raised both the quality and quantity of regulatory capital required to back risky assets. In addition, the introduction of the leverage ratio (typically 3%, but currently 5% for systemically important banks in the United States), which is calculated as a measure of capital relative to total exposures, including both on- and off-balance sheet items, also created a disincentive for banks to participate in market-making activities (Adrian, Boyarchenko and Shachar, 2017[33]).
Moreover, the framework introduced new liquidity standards to ensure that banks could meet both their short- and long-term funding needs. The treatment of corporate bonds under risk-weighting rules further reduced banks’ incentives to hold such assets. For instance, the Liquidity Coverage Ratio aims to ensure that banks have a 30‑day supply of cash and HQLAs to account for possible stressed scenarios (Morgan Stanley, 2018[34]). However, corporate bonds face a 15% haircut in the HQLA calculation, meaning that only 85% of their market value is accounted for in this measure.
In the United States, the Volcker Rule, introduced as part of the Dodd-Frank Act, also impacted banks’ behaviour by prohibiting proprietary trading. Proprietary trading differs from market-making in that the former involves banks using their own capital to profit from price movements, whereas the latter focusses on providing market liquidity and meeting client trading needs. Although market-making activities are exempt from the Volcker Rule, the regulation has still discouraged banks from holding large inventories of assets that are difficult to liquidate, such as corporate bonds, as maintaining large positions could raise regulatory scrutiny (Maureen and Zhou, 2025[35]).
These regulatory constraints have ushered in a shift in the creditor landscape, driving the growth of non-bank financing institutions (NBFIs), notably investment funds, which now account for roughly half of global financial assets (FSB, 2024[36]; OECD, 2024[30]). Combined with technological developments and the rise of electronic trading (see Chapter 2 of this report), they have also contributed to a transformation in the functioning of secondary bond markets. Banks’ traditional role as the primary source of liquidity is gradually shifting towards automated, transaction-based market-making. As a result, corporate bond liquidity is now provided in a more fragmented manner, and banks have fewer incentives to hold inventories of corporate bonds on their balance sheets than in the past. Meanwhile, in government bond markets, hedge funds, which are not subject to the same regulations as banks, have become the marginal providers of liquidity, often accounting for a majority of secondary market trading volumes (Box 3.3).
Changes to the enhanced supplementary leverage ratio (eSLR) in the United States could provide more capacity for many of the world’s largest banks to engage in low-risk activities such as intermediation in US Treasuries and repo financing. The rule change would lower the requirement on Global Systemically Important Banks (G-SIBs) and their subsidiaries by replacing the 2% buffer (which sits on top of the 3% minimum supplementary leverage ratio) with a buffer equal to 50% of each G-SIB’s Method‑1 risk surcharge, capped at 1% for subsidiaries.5 This is expected to free up roughly USD 13 billion in Tier‑1 capital at the holding company level and approximately USD 219 billion at major bank subsidiaries (FDIC, 2025[37]). The change alone will not necessarily encourage banks to buy Treasuries on a large scale, but the reform should be, all things being equal, liquidity-enhancing for the US Treasury market.
One of the consequences of the post-global financial crisis banking regulations has been an increase in the costs associated with conducting primary dealer activities. Due to increased capital requirements and tighter regulations, banks have scaled down both their warehousing function and proprietary trading activities. Equally, because of balance sheet constraints, banks’ trading desks are less flexible in digesting large supply and demand imbalances of government bonds (BIS, 2014[38]). In recent years, hedge funds have increasingly filled the gap left by banks, profiting from small price discrepancies by using leveraged trades. While banks reduced their direct engagement in government bond markets, they enable hedge fund activities by providing leverage through repurchase operations (repo).
In many jurisdictions, hedge funds, which are subject to different regulatory regimes, have become the marginal providers of liquidity. By early 2025, they accounted for nearly a third of secondary market trading volumes in US Treasuries (Figure 3.15, Panel A) and the majority in sovereign bond markets in the United Kingdom and the Euro area (Figure 3.15, Panels B and C) on the trading platform Tradeweb.
Note: Data as of February 2025.
Source: Tradeweb via Stefan Goghie (2025[39]), Hedge funds and government bond markets: Why is there a love story? https://goghieas.substack.com/p/hedge-funds-and-government-bond-markets.
DMOs have also reported a growing role for “other market participants” in activities like relative‑value (RV) trading or warehousing of bonds. Eleven OECD DMOs have observed this in the last 12 months, with Belgium, Finland, Lithuania and Spain recognising hedge funds specifically as playing a bigger role. In Japan, hedge funds were reportedly increasingly entering the JGB market due to various factors, including the steepening of the yield curve. In the Netherlands, hedge funds are seen as being more active since the end of the ECB’s QE programme in 2023. Meanwhile, more than 60% of surveyed DMOs reported hedge funds engaging in RV trading (Figure 3.16, Panel A), a role traditionally undertaken by banks.
In addition to supporting secondary market liquidity, many DMOs have also reported that hedge funds provide a direct source of demand, supporting supply events. Belgium reported that hedge funds buy roughly 40% of the supply on auction days. At Canadian Government bond auctions, hedge fund bidding has strongly increased in recent years. The share of hedge fund bids relative to total bids increased from single‑digit percentages in 2010-2015 to an average share of about 30% in recent years, at relatively competitive levels, resulting in allocations of above 40% (Epp and Gao, 2025[6]). In Germany, where hedge funds cannot directly participate at auctions, they have become the largest net buyer of auctioned securities in the secondary market on auction days (Branger, Muck and Pütz, 2024[40]).
Further, more than half of surveyed OECD DMOs reported that hedge funds have been marginal buyers of their bonds in the last 12 months (Figure 3.16, Panel A). This means that as the net supply of bonds to the market has increased, they have increased their purchases by a greater proportion. Italy saw hedge funds become marginal buyers for the first time during that period. In total, 11 DMOs reported either moderate or significant inflows from hedge funds into their bond markets in the last 12 months, with only one reporting net outflows.
Measuring the precise holdings by hedge funds is challenging due to their domicile, holding structures, and more general difficulties in identifying the beneficial holders of bonds. Most DMOs do not have access to the necessary data and instead rely on feedback from market participants. On this basis, Canada estimates that hedge funds were allocated around 40% of its bond issuance in 2025. Where this data is collected and been shared with the OECD, the trend is clear since 2022. The approximate share of hedge fund holdings of Mexican Government bonds has risen from 14% in 2022 to 17% in 2025. In the US, their share has almost doubled from 5% to 9% over the same period, whilst in Germany, hedge fund net holdings have risen from ‑3% (due to short selling) to 2% over the same period (Figure 3.16, Panel B).
Note: Panel A is based on responses by 37 DMOs. In Panel B, the data for Germany shows net holdings, which are negative in 2022 due to short selling.
Source: 2025 OECD survey on secondary market liquidity in government bond markets; Deutsche Finanzagentur; and OECD calculations; BIS (2014[38]): “Market-making and proprietary trading: industry trends, drivers and policy implications”, CGFS Papers No 52, https://www.bis.org/publ/cgfs52.htm.
As Box 3.2 shows, hedge funds were net buyers of about EUR 75 billion of German Government bonds between June 2022 and November 2025. Whilst hedge funds are the least likely group to hold bonds to maturity, they are therefore not likely to become net holders on the same scale as central banks or foreign investors. However, it is clear they are playing a growing, in some cases vital role in providing liquidity, and have become the marginal buyers in many jurisdictions.
The risks from greater reliance on hedge funds are clear: higher failure rates, greater flight risk, and higher refinancing risk on margin or leverage calls; in addition to lower transparency compared to traditional market makers (banks), and end investors (pension and insurance funds); and less comprehensive and homogenous regulatory oversight. Despite this, only one DMO reported that the involvement of hedge funds negatively impacts liquidity in periods of heightened market stress, and none reported a decline in normal circumstances. However, four DMOs reported that hedge funds increase market liquidity in both scenarios.
Insurance company regulation
Capital adequacy frameworks such as the Risk-Based Capital (RBC) system in the United States and the Solvency II framework in Europe have increasingly influenced how insurance companies decide on their bond allocations. These frameworks link required capital to risk exposure, ensuring solvency under stress but also constraining risk-taking. Since the global financial crisis, the tightening of such regulations has created incentives for insurers to limit their exposure to corporate bonds, particularly those with longer maturities or lower credit ratings. Under Solvency II, for instance, assets are valued at market prices, so fluctuations in bond values directly affect the volatility of insurers’ balance sheets, discouraging large holdings of corporate bonds. Moreover, insurers must hold capital against spread risk, with charges increasing with bond maturity, which further reduces the attractiveness of long-term bonds. In contrast, EU sovereign bonds are exempt from capital charges, regardless of their rating or maturity (European Commission, 2017[41]). Similarly, in the United States, US Government bonds are excluded from risk-based capital requirements (NAIC, 2021[42]).
Furthermore, since rating-based capital requirements, such as those under the RBC framework in the United States and Solvency II in Europe, assign different capital charges depending on credit ratings, there is evidence that insurers often avoid bonds near the lower end of the investment-grade spectrum (e.g. BBB) to mitigate the risk of higher capital charges and potential losses in the event of a downgrade to non-investment grade. There is also evidence that increasing regulatory constraints on insurance companies are associated with greater tendencies by the companies to sell downgraded bonds (Ellul, Jotikasthira and Lundblad, 2011[43]). Within the same capital charge category, however, insurers appear to favour bonds with higher systematic risk (not captured by credit ratings) to enhance expected returns (Murray and Nikolova, 2021[44]), reflecting the search-for-yield behaviour commonly observed in insurance companies (Becker and Ivashina, 2014[45]). This could, for example, lead to a concentration of holdings in certain industries.
Pension funds’ specific motives for investing in bonds differ significantly depending on whether they operate under a DB or DC structure. DB schemes, similar to life insurance companies, promise a fixed payout to policyholders, which requires them to closely match the duration of their assets and liabilities. They are also subject to statutory minimum funding ratios to ensure that future payments can be met. These factors encourage investment in long-term securities, particularly government bonds (ECB, 2021[32]). In contrast, DC schemes place the investment and longevity risk on individual policyholders. Since they have no fixed liability to hedge, these schemes prioritise higher expected returns and typically hold a larger share of equities and investment fund units.
Source: DNB (De Nederlandsche Bank).
Over recent decades, many pension systems have shifted from DB to DC schemes. This transition has been driven by interrelated factors, including changes in legislation, regulation, and tax policy that made DB plans more complex and costly to administer. Ageing workforces and higher life expectancy have increased DB funding costs, while rising labour mobility has encouraged workers to opt for DC plans, which allow benefits to be transferred or accumulated without penalties. In the United States, the introduction of 401(k) has also accelerated the adoption of DC schemes (Broadbent, Palumbo and Woodman, 2006[46]).
This gradual transition has clear implications for bond allocations. As the share of DC plans rises, and in the absence of any changes to the regulation of these schemes, aggregate demand for ultra-long bonds is expected to decline, and portfolios are expected to be increasingly weighted towards equities and investment fund shares. There could also be a shift from government to corporate bonds, given the higher yields. Notably, Dutch pension funds, the largest in Europe, began shifting from DB to DC schemes in 2023, a process expected to conclude by 2027. This transition is expected to raise the share of DC schemes in the Euro area from around 17% to 77% (ECB, 2021[32]), which could significantly influence asset allocation across the region.
The impact of this shift in the holding structure of Dutch pension funds is already beginning to materialise. Allocations by Dutch pension funds to government and corporate debt securities already fell below 50% in 2021 for the first time and are likely to continue to fall (Figure 3.17, Panel A). This is also resulting in a slight shift away from government bonds and towards corporate bonds (Figure 3.17, Panel B).
Equally, as DB schemes close to new members, with many members approaching or reaching retirement and therefore beginning to draw their pensions, remaining demand is moving down the maturity curve. This is already evident in the holding structure: holdings of 20+ year government bonds have declined consistently since 2020 (Figure 3.17, Panel C), and holdings of 20+ year corporate bonds have reached their lowest levels since before the pandemic (Figure 3.17, Panel D). Again, both are expected to decline further.
Note: In Panel C, the UK DMO defines long-dated bonds with a maturity of 15 years or more; 2025 data are expectations, and 2026 data are projections.
Source: OECD; Office for Budget Responsibility (OBR); UK DMO.
Also in the United Kingdom, where DB pension schemes have been an important source of demand for long-dated and inflation-linked UK Government bonds, helping the UK to country maintain the longest ATM in its government debt portfolio in the G7 (Figure 3.18, Panel B), the move to DC pensions and the “winding down” of DB pensions has resulted in a significant change in issuance plans, with a far smaller proportion of long-dated and inflation-linked bonds being issued in recent years (Figure 3.18, Panel C). In addition, it is expected that most remaining DB pension schemes will move towards a “buy-out” transaction with an insurance fund. These funds can be slightly more yield sensitive than pension funds and are more likely to consider the attractiveness of bond valuations on a relative or asset swap basis. They are also subject to slightly different regulatory regimes, meaning they tend to gravitate towards higher-yielding assets.
As shown in the previous sections, the level and shape of the yield curve can impact the composition of the investor base. Importantly, this relation runs in both directions. The following section examines how changes in the holding structure feed back into market pricing, liquidity, and volatility, as well as into issuers’ behaviour.
The relevance of the investor base for bond valuation
In frictionless financial markets with unconstrained arbitrage, asset prices should only reflect their fundamental value, irrespective of who holds them. The experience of recent decades has, however, provided compelling evidence to the contrary. It is now clear that shifts in the investor base can exert a strong influence on prices. Consequently, bond yields can at times deviate significantly from their fundamental values.
Accordingly, QE programmes aimed to lower yields by reducing the free float of outstanding securities with the precise impact depending on issuance mix, repo market conditions, and demand for safe assets. Empirical evidence confirms that this policy has been successful in influencing asset prices. QE not only reduced yields in the directly affected segments of the yield curve (see e.g. D’Amico and King (2013[47]), Froemel, Joyce, and Kaminska (2022[48])), but also caused a downward shift of the entire curve (Eser et al., 2019[49]). This drove down not just government bond yields but also those of corporate bonds (Krishnamurthy and Vissing-Jorgensen, 2011[50]). These effects were most pronounced in markets with low levels of bond issuance, as was the case in Euro area countries with relatively low government debt between 2015 and 2020, where bond yields fell significantly below counterfactual estimates. In addition, bond-swap spreads and repo rates indicated scarcity, as price‑insensitive investors in need of high-quality collateral and safe assets pushed prices to abnormal levels (Baltzer, Schlepper and Speck, 2025[51]).6
However, monetary policy is not the only factor influencing bond yields. Structural changes in the behaviour of important investor groups can also have significant effects on the shape of yield curves, as illustrated by the following recent examples. One is the migration of Dutch pension funds from DB to DC schemes (see Box 3.4), regarded as one of the main drivers of the steepening of Euro area yield curves over the past year (Bloomberg, 2025[52]). Another is the pro-cyclical selling of government bonds by liability-driven investors in the United Kingdom in 2022, which was a major factor behind the surge in long-end yields (Pinter, 2023[53]).
Price‑sensitive investors play a crucial role in helping the broader market absorb demand-supply imbalances, thereby smoothing price developments over time and across securities. Consequently, their increasing presence usually reflects pronounced supply increases and/or reduced structural demand by price‑insensitive investors. The share of domestic investors classified as price‑sensitive in government bonds markedly decreased between 2013 and 2021 (Figure 3.19, Panel A).7 With the end of QE and the high issuance of the last years, it has since then increased by around 15 p.p. and now accounts for more than half of domestic absorption. A similar trend can be observed for foreign holders, where the share of the less price‑sensitive official sector has markedly decreased since 2018 (Figure 3.19, Panel B).
More price‑sensitive investors now account for a majority of domestic holdings of government bonds, while the share of foreign holders accounted for by non-banks has been rising since 2020
Note: In Panel A, data refer to general government and cover the Euro area, Japan and the United States. Yield insensitive investors refer to the central bank, insurance companies and pension funds. Price‑insensitive investors include all other categories. Panel B covers Japan, the United States, and all Euro area countries except Croatia, Cyprus, Estonia, Latvia, Lithuania, Luxembourg, Malta and the Slovak Republic. Data for Panel B are available until 2024.
Source: Panel A: Bank of Japan; European Central Bank; US Federal Reserve; Panel B: International Monetary Fund: Sovereign Debt Investor Base for Advanced Economies.
How the investor base affects liquidity and volatility
One of the key characteristics of government bonds in OECD countries is their high liquidity, which allows market participants to execute large transactions swiftly and at minimal trading cost. As government bond prices largely reflect discounted cash flows, bond market volatility is usually primarily driven by changes in market expectations regarding future monetary policy and risk perception. This differs from corporate bonds, where outstanding amounts of single bond lines are usually much smaller, and trading costs are higher (although this is partially changing – see Chapter 2).
However, the holding structure of bonds can have significant effects on both liquidity and volatility even in large bond market segments. A strong presence of price‑insensitive investors – such as central banks, insurance companies, and pension funds – can result in a low free float of bonds readily available for purchase. The resulting scarcity of bonds drives up their prices and increases their bid-ask spreads. In the most extreme events, this can lead to so-called short squeezes characterised by a high relative value of the respective instruments and sharply falling repo rates, indicating the high costs of even temporarily borrowing the respective security (Baltzer, Schlepper and Speck, 2025[51]). In contrast, in corporate bond markets, empirical evidence suggests that central bank purchases can have a liquidity-enhancing effect (Boneva et al., 2022[54]).
Surges in bond supply, on the other hand, can trigger elevated volatility and increased transaction costs when the capacity of price‑sensitive actors, such as banks’ trading desks, is constrained. This was particularly evident during the “dash for cash” episode in March 2020, when a broad range of investors sought to liquidate holdings of long-duration instruments. The result was not only a sharp rise in yields but also a pronounced widening of bid-offer spreads, even in the most liquid government bond markets (Barone et al., 2022[55]).
Rising net supply implies greater reliance on hedge funds, MMFs, and other balance‑sheet-constrained intermediaries, increasing sensitivity to repo capacity and funding conditions. Under normal market conditions, the growing presence of hedge funds in bond markets tends to enhance liquidity and smooth price developments. However, it can also amplify market dysfunction, and liquidity can wane when conditions become strained. By nature, hedge funds are lightly regulated and typically operate with high leverage. Unlike market makers, they do not necessarily aim to provide a reliable source of liquidity to their trading partners. Consequently, during periods of turbulence, hedge funds tend to liquidate positions and withdraw from the market precisely when liquidity support is most needed.
A prominent example of such an event was the “Liberation Day” yield spike in April 2025, when hedge funds collectively liquidated positions in US Government bonds against interest rate swaps (Financial Times, 2025[56]). With the significant share of bonds currently held by leveraged investors, there is an elevated risk that a large‑scale unwind of these positions could lead to a sharp and self-reinforcing selloff in bond markets.
In times of market turbulence, central banks can play a backstop function as a buyer of last resort by reducing the amount of duration and credit risk the market has to absorb, or as a market maker of last resort, easing sudden supply-demand imbalances impairing market functioning (Guttmann, 2012[57]; Kashyap et al., 2025[58]). While central banks have the technical firepower to address market turbulences by taking the lender of last resort roles, this can be regarded as problematic at least in two aspects. First, the expectation of central bank interventions might create moral hazard among market participants, resulting in higher risk taking. Second, the aim to ensure market functioning can’t interfere with the monetary policy mandate of central banks.
Therefore, while the increased presence of highly leveraged, speculative participants might be welcomed in normal times due to their liquidity provision, it also creates risks for market functioning and financial stability during periods of market stress. Future market resilience does not only hinge on investor shares, but even more importantly, on who absorbs marginal supply, under which policy and regulatory constraints, and at what price. Depending on this, absorption capacity is therefore likely becoming more conditional and fragmented.
The interplay between government and corporate bond markets
A key link between the investor base and market pricing is given by the interplay between government and corporate bond markets. As government bonds in most economies serve as pricing benchmarks for corporate bonds, developments in government bond markets have profound implications for corporate markets. One such dynamic in the post-2008 period was lower yields in government bond markets pushing investors into longer-dated government bonds or riskier corporate bonds in search of higher yields. This in turn impacts liquidity and bond valuation.
“Yield chasing” or “reaching for yield” describes the dynamic in which investors increase their exposure to riskier assets to obtain higher yields when general interest rates are low. This can be achieved in the bond market by taking on greater duration risk, credit risk, or capturing illiquidity premia. In the first case, investors purchase longer-term bonds, which typically offer higher yields – driven by the term premium – thereby extending portfolio duration and increasing sensitivity to interest rate movements. Alternatively, they may invest in lower-rated bonds that provide higher credit spreads as compensation for elevated default risk, or less liquid assets.
This behaviour has been documented across a wide range of investor types, including mutual bond funds (Choi and Kronlud, 2017[59]), commercial banks (Hanson and Stein, 2015[60]), hedge funds (Czech and Robert-Sklar, 2019[61]), individual investors (Huang et al., 2025[62]; Gomes et al., 2023[63]), insurance companies (Becker and Ivashina, 2014[45]; Czech and Robert-Sklar, 2019[61]), MMFs (Maggio and Kacperczyk, 2017[64]) and pension funds (Andonov, Bauer and Cremers, 2017[65]). Dealers tend to exhibit the opposite behaviour though – selling riskier bonds when interest rates are lower – facilitating the “yield chasing” of other investors (Czech and Robert-Sklar, 2019[61]).
On the international level, there is evidence that foreign investors tend to shift their bond portfolios towards riskier US corporate bonds when interest rates in their home countries decline (Ammer et al., 2018[66]). A similar shift to riskier US bonds has been found for foreign investment funds in response to US monetary policy tightening, as it raises the currency-hedging costs of holding US dollar assets (Ahmed, Hofmann and Schmitz, 2023[67]).
The drivers of yield chasing vary by investor type. For individual investors, decisions are mainly influenced by personal preferences and psychological factors, reflecting differences in how people assess the trade‑off between risk and return in different interest rate environments (Lian, Ma and Wang, 2018[68]). Since mutual funds cater to these investors, these same motivations apply to them as well (Choi and Kronlud, 2017[59]). In contrast, for insurance companies and US public pension funds, reaching for yield often serves as a form of regulatory arbitrage. Insurance companies are subject to capital requirements that depend on the riskiness of their assets, which encourages them to take on additional risk for a given rating (Becker and Ivashina, 2014[45]).
Public pension funds in the United States, which can set their liability discount rate based on expected asset returns, are motivated to reach for yield to lower the present value of liabilities and present a stronger funding position (Andonov, Bauer and Cremers, 2017[65]). For banks, accounting rules that require earnings to be reported based on current income from securities incentivise them to pursue reach-for-yield behaviour (Hanson and Stein, 2015[60]). However, when it comes to increasing duration risk among insurers in a low-rate environment, evidence suggests this may not primarily serve to increase yields, but rather to address the widening asset-liability duration gap that follows from “negative convexity”, i.e. the fact that given a duration mismatch, the value of insurance liabilities rises more than the value of assets when rates fall (Domanski, Shin and Sushko, 2015[15]).
Especially in times of weak economic activity, reaching for yield can promote economic growth by lowering the financing costs for riskier borrowers and easing overall financial conditions, which is a key purpose of expansionary monetary policy in general and QE in particular. It may, on the other hand, also generate adverse effects on financial stability through mispricing of risk in credit markets and other assets (Becker and Ivashina, 2014[45]; Choi and Kronlud, 2017[59]; Chen and Choi, 2022[69]). Moreover, reaching for yield can have an outsized effect on the long end of the yield curve by disproportionately increasing demand for longer-maturity bonds, thereby compressing long-term yields and flattening the yield curve (Hanson and Stein, 2015[60]).
The impact of the changing investor base on issuance strategies
The reduced demand for duration seen in recent years has implications for issuers’ strategies. As discussed in Chapter 1, several government issuers have skewed their issuance plans shorter – issuing a greater share of short and medium-dated bonds, made greater use of syndications, tapped more “off the run” bonds, and made great use of T-bills in response to this changed demand profile.
However, as governments are typically repeat borrowers and look to pursue predictable and transparent issuance plans, whilst looking to support liquidity at all issuance points across their yield curves, these changes are typically incremental. Corporates tend to have more flexibility, not only to change the maturity of issuance but to reduce borrowing in response to higher yields and therefore borrowing costs. The share of issuance of >20‑year US Treasuries and >20‑year US investment grade corporate bonds, plotted against the benchmark 30‑year US Treasury yield shows this (Figure 3.20). This suggests that corporate issuers are more nimble and able to react to changes in long-term yields more quickly than a large government issuer, changing their issuance splits accordingly. This opportunism can have implications for market signals, though, notably reducing the viability of secondary market spreads as a predictor of primary market spreads (Boyarchenko and Elias, 2024[70]).
Corporate issuers have been more responsive to changes in long-term bond yields than sovereigns
Source: Bloomberg; LSEG; OECD calculations.
From a debt management perspective, shortening the maturity of issuance increases refinancing and refixing risk. From a market perspective, it concentrates issuance in the typically more liquid parts of the curve (T-bills, 2‑year, 5‑year and 10‑year maturities). Such an issuance profile tends to attract more activity from certain investor types such as MMFs, retail investors, sovereign wealth funds, and bank treasuries who either prefer to buy shorter maturities or have mandates that prevent them from trading beyond a certain point on the curve. In addition, the greater liquidity in these parts of the curve provides more opportunities for arbitrage and RV trades, which are now largely conducted by hedge funds. Hence, it is possible that issuance strategies will help to crystallise the trends observed in investor bases in the future.
Looking ahead, several factors could impact the investor base for government and corporate bonds. Prime amongst these are the continued shift from DB to DC pension schemes, the persistence of higher yields, particularly at longer maturities, the growing role of hedge funds, evolving geopolitical tensions, and developments in monetary policy. New technologies are also impacting bond markets, and these may reallocate demand or even result in new demand in certain markets (see Box 3.5).
The shift from DB to DC pension schemes will likely mean an overall reduction in demand for bonds in favour of equities and alternative investments by this sector, but also a shift within bond holdings from government to corporate, and from longer to shorter-term. DB pension schemes will remain less prevalent, with the vast majority now closed to new members, whilst their current members are approaching or have reached retirement. With the share of assets in DC plans continuing to rise, the pension sector as a whole may make an allocation shift out of bonds, particularly from lower-yielding government and high-grade corporate bonds (ECB, 2021[32]).
The change in demand associated with this shift is most relevant for some advanced European markets (particularly the Netherlands and the United Kingdom), where DB funds historically provided stable, long-duration demand for government and inflation-linked bonds. In other regions, notably the United States, this transition occurred earlier and is largely already reflected in portfolio structures, while in Japan, the more dominant forces are demographics and institutional design rather than DB-to-DC reform. More importantly, the shift implies less a disappearance of pension demand, and more a change in its nature: from liability-anchored, price‑insensitive demand towards more valuation- and cycle‑sensitive allocations, often implemented via funds or derivatives rather than cash bonds.
Equally, the persistence of higher yields could have mixed effects on demand from insurers and asset managers. Insurers could be less incentivised to invest in riskier assets to achieve an attractive spread revenue,8 as they were during the extremely low-interest rate environment pre‑2022. However, higher yields can be equity-positive for life insurers because liability duration is higher than asset duration, increasing risk-taking in a higher-rate environment (Li, 2025). Moreover, the higher-yield environment is favourable for asset managers investing in bonds. This is because they provide a greater cushion against adverse moves in rates (rising) and compound on top of price appreciation in an advantageous scenario (falling rates).
Hedge funds’ continued and growing presence in sovereign bond markets may continue to provide much needed liquidity and help to smooth out mispricing, but it could also render the investor base riskier due to flight risk and pressure on repo markets. Hedge funds pose a relatively high flight risk given that they have a high failure rate globally, they face higher refinancing risks on leverage or margin calls, and they lack organic commitment to foreign markets. The rising activity of hedge funds could also place pressure on the broader fixed-income infrastructure because their market positioning is financed through the repo market, raising questions of repo market capacity.
Concerns related to geopolitical risk remain prominent in both the sovereign and corporate bond markets
Note: In panel A total surveyed countries were 41 (2022), 41(2023), 38(2024), 35(2025). In 2025, answers were given in October. Panel B refers to companies in the Russell 3 000, euro STOXX 600 and NIKKEI 225 indices. Mentions in corporate filings and earnings call transcripts.
Source: 2022, 2023, 2024, and 2025 OECD Surveys on Liquidity in Government Bond Secondary Markets; Corporate document analysis via Bloomberg, see Annex 3.A for details.
Geopolitical risk is a factor to monitor closely in sovereign and corporate bond markets. Nearly all surveyed sovereign issuers expect geopolitical risk to affect primary market operations in 2026 (Figure 3.21, Panel A). Similarly, corporate disclosure – financial and non-financial filings and earnings call transcripts – suggests that concerns about geopolitical risk remain elevated (Figure 3.21, Panel B). Beyond market volatility, geopolitical risk could have meaningful implications for the composition of the investor base.
Stablecoins are digital tokens designed to maintain a stable value relative to a certain asset. They typically claim to be backed one‑to‑one by high-quality, liquid assets such as short-term government securities (T-bills) held in reserve, helping to stabilise their price. Some market participants argue that the increasing adoption of stablecoins represents a potential source of increased, mostly price‑insensitive demand for T-bills. This view rests on the assumption that stablecoin issuers can increase net demand, i.e. that it does not crowd out existing demand for T-bills.
Stablecoin issuers may increase the gross demand for T-bills, but the key aspect is whether the additive or substitutive effect dominates. For example, selling MMF shares to buy stablecoins should have a largely neutral impact on the demand for T-bills. This is because the MMF sells its holdings, while the stablecoin issuer buys a proportional amount of T-bills as the underlying reserve. However, switching from physical cash or bank deposits to stablecoins should lead to a decline in currency in circulation or deposits and an increase in demand for T-bills. Despite this, demand for longer-term government bonds could decrease if banks fund deposit outflows by selling these securities. Demand for T-bills will also rise if stablecoin issuers receive inflows from foreign investors. The rate environment also matters. In a scenario of a return to an ultra-low-rate environment, stablecoin issuers may shift to higher-yielding alternatives for high-liquid reserves, which would also have implications for T-bill demand. This will ultimately also depend on the regulatory framework and what it prescribes as possible reserve assets.
However, the limited transparency and inconsistent disclosure of stablecoin reserve compositions constrain visibility into their actual asset holdings. As a result, it is difficult to assess conclusively whether stablecoins can generate new net demand for short-term government securities or simply reallocate demand from existing holders such as MMFs or banks.
Note: Data and projections as of March 2025.
Source: A: Aldasoro, Iñaki and Rashad Ahmed (2025[71]): “Stablecoins and safe asset prices”, BIS Working Paper 1 270, US Treasury.
Currently, the stablecoin market is dominated by USD-based products; they accounted for 99% of total market capitalisation in 2024 (Schaaf, 2025[72]). Based on publicly available reserve filings, major stablecoin issuers were estimated to hold more than USD 120 billion in US T-bills as of March 2025, higher than holdings by investors in several key geographies, including Luxembourg, the United Kingdom, and China (Figure 3.22, Panel A). While this figure represents approximately 2% of outstanding US T-bills, the US Treasury estimates that the rapid growth could generate a substantial increase in demand, with incremental purchases potentially reaching US dollars 900 billion by 2028 (Figure 3.22, Panel B) (US Treasury, 2025[73]). However, the limited transparency and inconsistent disclosure of stablecoin reserve compositions constrain visibility into their actual asset holdings. As a result, it is difficult to assess conclusively whether stablecoins can generate new net demand for short-term government securities or simply reallocate demand from existing holders such as MMFs or banks.
Geopolitical tensions and uncertainty can lead to a shift in portfolio allocations by impacting investors’ expectations and willingness to tolerate risk. In 2025, almost all OECD sovereign issuers that reported a significant change in their investor base identified geopolitics as a factor (Figure 3.23, Panel A). Two main dynamics could be driving this. First, investors may shift their portfolio towards bonds with shorter maturities. Second, they may shift to bonds perceived as safe‑haven assets, explaining why the US Government bond market attracts funds during times of heightened geopolitical risk (Figure 3.23, Panel B) (Converse and Mallucci, 2025[74]). These reallocations can result in higher term premia and higher spreads between “safe haven” bonds and others (Constantini and Sousa, 2022[75]; Papavassiliou, 2025[76]).
Sovereign issuers assess that geopolitical factors are one of the main drivers of changes in the investor base
Note: Answers shown in panel A were given in October 2025 and refer to the previous 12 months. In Panel B, the yearly geopolitical risk (GPR) index is calculated as the average monthly index.
Source: 2025 OECD Surveys on Liquidity in Government Bond Secondary Markets, US Federal Reserve; Caldara and Iacoviello (2022[77]), Measuring Geopolitical Risk, https://doi.org/10.1257/aer.20191823.
Geopolitical tensions can especially impact demand from foreign investors. An escalation in geopolitical risk often contributes to a contraction of cross-border capital flows, generating a “flight home” effect, whereby investors reallocate funds towards domestic markets (Feng et al., 2023[78]; Catalán and Tsuruga, 2023[79]; Agoraki et al., 2024[80]). In addition, geopolitical alignment increasingly shapes the direction of international capital. For example, investment funds tend to allocate smaller shares of their portfolios to countries that are geopolitically more distant (Catalan, Fendoglu and Tsuruga, 2024[81]). Early evidence also indicates that geopolitical distance may affect foreign official holdings of sovereign debt (Beck et al., 2025[82]). Such cross-border reallocations of capital can reduce the number of available financial partners for some countries, heightening their vulnerability to external shocks by limiting their capacity for risk diversification (IMF, 2023[83]).
The potential impact on foreign investor demand is particularly notable, as recent estimates indicate that roughly one‑fifth of the US sovereign debt and one‑quarter of the European sovereign debt were held by geopolitically non-aligned countries in 2024 (Beck et al., 2025[82]). Rising geopolitical tensions may also dampen domestic investor sentiment, with evidence suggesting that heightened uncertainty weakens consumer confidence and reduces households’ willingness to invest (Coibion et al., 2025[84]). This, in turn, could curb domestic demand for bonds, particularly among households and investment funds, whose reluctance to invest in countries perceived as geopolitically risky tends to persist over time (Converse and Mallucci, 2025[74]).
Geopolitical risk can also affect borrowing costs, market liquidity, and issuance strategies. Recent estimates suggest that heightened geopolitical uncertainty increases investors’ perception of sovereign risk, thereby putting upward pressure on government bond yields (Afonso, Alves and Monteiro, 2024[85]). In 2024, nearly half of the countries surveyed by the OECD reported that geopolitical tensions had adversely affected liquidity conditions in domestic bond markets. Consistent with this, sovereign issuers have identified geopolitical risk as one of the primary factors affecting their operations in the primary market every year since 2022. Elevated geopolitical uncertainty can also alter corporate financing behaviour, prompting firms to shorten the maturity of their bond issuances or shift away from bond markets towards bank lending (Khoo and Cheung, 2021[86]).
From a structural perspective, geopolitical tensions that influence trade policy can generate sustained changes in current account balances. If shifts in the geopolitical landscape lead to durable reductions in external surpluses, the corresponding decline in deficits would reduce an important source of demand for the debt issued by deficit countries and their corporations. While lower deficits are, by definition, matched by reduced surpluses, the resulting contraction in available foreign financing may constrain the ability of deficit economies to roll over existing debt, potentially leading to forced deleveraging.
Finally, differences in central bank operating frameworks and balance‑sheet trajectories will be critical. The US Fed’s transition towards so-called reserve‑management purchases, which will see the balance sheet expand again in the first quarter of 2026, the ECB’s continuation of QT within a corridor system, and the BoJ’s gradual and asymmetric exit from bond markets imply structurally different paths for free float, term premia, and marginal demand. These distinctions dictate that future investor-base dynamics are not only about portfolio preferences, but also about how central bank frameworks shape net supply and collateral conditions over time.
References
[11] ABP (2025), Overview corporate bonds, https://www.abp.nl/content/dam/abp/documenten/beleggen/abp-overview-corporate-bonds.pdf.
[12] ABP (2025), Overview government bonds, https://www.abp.nl/content/dam/abp/documenten/beleggen/abp-overview-government-bonds.pdf.
[33] Adrian, T., N. Boyarchenko and O. Shachar (2017), “Dealer Balance Sheets and Bond Liquidity Provision”, Journal of Monetary Economics, Vol. 89, pp. 92-109, https://doi.org/10.1016/j.jmoneco.2017.03.011.
[85] Afonso, A., J. Alves and S. Monteiro (2024), “Beyond borders: Assessing the influence of Geopolitical tensions on sovereign risk dynamics”, European Journal of Political Economy, Vol. 83, https://doi.org/10.1016/j.ejpoleco.2024.102550.
[80] Agoraki, M. et al. (2024), “Money never sleeps: Capital flows under global risk and uncertainty”, Journal of International Money and Finance, Vol. 141, https://doi.org/10.1016/j.jimonfin.2023.103013.
[67] Ahmed, A., B. Hofmann and M. Schmitz (2023), “Foreign institutional investors, monetary policy, and reaching for yield”, BIS Working Papers No. 1153, https://www.bis.org/publ/work1153.htm.
[71] Ahmed, R. and I. Aldasoro (2025), “Stablecoins and safe asset prices”, BIS Working Papers No. 1270, https://www.bis.org/publ/work1270.pdf.
[91] Aldridge, P., J. Sandhu and S. Tchamova (2024), “How foreign central banks can affect liquidity in the Government of Canada bond market”, Staff Analytical Note, https://doi.org/10.34989/san-2024-26.
[66] Ammer, J. et al. (2018), “Searching for yield abroad: risk-taking through foreign investment in US bonds”, BIS Working Papers No. 687, https://www.bis.org/publ/work687.pdf.
[65] Andonov, A., R. Bauer and K. Cremers (2017), “Pension fund asset allocation and liability discount rates”, The Review of Financial Studies, Vol. 30/8, pp. 2555-2595, https://doi.org/10.1093/rfs/hhx020.
[5] Aquilina, M., G. Cornelli and N. Tarashev (2025), Commonality under pressure: banks and funds, BIS Quarterly Review, https://www.bis.org/publ/qtrpdf/r_qt2503e.htm.
[51] Baltzer, M., K. Schlepper and C. Speck (2025), “The Eurosystem’s asset purchase programmes, securities lending and Bund specialness”, Journal of Business Economics, https://doi.org/10.1007/s11573-025-01243-w.
[26] Bank of Japan (2025), Minutes of the Monetary Policy Meeting on June 16 and 17, 2025, https://www.boj.or.jp/en/mopo/mpmsche_minu/minu_2025/g250617.pdf.
[55] Barone, J. et al. (2022), “The Global Dash for Cash: Why Sovereign Bond Market Functioning Varied across Jurisdictions in March 2020”, New York Fed Staff Reports No. 1010, https://www.newyorkfed.org/medialibrary/media/research/staff_reports/sr1010.pdf?sc_lang=en.
[87] Barth, D. et al. (2025), “The Cross-Border Trail of the Treasury Basis Trade”, FEDS Notes, https://doi.org/10.17016/2380-7172.3939.
[45] Becker, B. and V. Ivashina (2014), “Reaching for yield in the bond market”, The Journal of Finance, Vol. 70/5, pp. 1863-1902, https://doi.org/10.1111/jofi.12199.
[82] Beck, R. et al. (2025), Geopolitics and foreign holdings of euro area government debt, European Central Bank, https://www.ecb.europa.eu/press/other-publications/ire/article/html/ecb.ireart202506_01~a8b7241329.en.html.
[89] Belz, S. and D. Wessel (2020), “What is yield curve control”, Brookings, https://www.brookings.edu/articles/what-is-yield-curve-control/.
[38] BIS (2014), “Market-making and proprietary trading: industry trends, drivers and policy implications”, CGFS Papers No. 52, https://www.bis.org/publ/cgfs52.htm.
[52] Bloomberg (2025), A €2 Trillion Dutch Pension Headache Is Coming for European Bonds, https://www.bloomberg.com/news/articles/2025-08-31/dutch-pension-reform-risks-turning-into-a-2-trillion-headache.
[10] Board of Governors of the Federal Reserve System (2025), Financial Accounts of the United States - Z.1, https://www.federalreserve.gov/releases/z1/default.htm.
[96] Boitreaud, S., S. Foxall and L. Policino (2025), “Retail investors and sovereign debt: Why now and what next?”, https://www.oecd.org/en/blogs/2025/05/retail-investors-and-sovereign-debt-why-now-and-what-next.html.
[54] Boneva, L. et al. (2022), “The Impact of Corporate QE on Liquidity: Evidence from the UK”, The Economic Journal, Vol. 132/648, https://doi.org/10.1093/ej/ueac033.
[70] Boyarchenko, N. and L. Elias (2024), Corporate Credit Conditions Around the World: Novel Facts Through Holistic Data, Federal Reserve Bank of New York, https://doi.org/10.59576/sr.1074.
[40] Branger, N., M. Muck and A. Pütz (2024), “Unmasking Global Investors: An Examination of the Shifting Dynamics in Sovereign Bond Trading”, https://doi.org/10.2139/ssrn.5057388.
[46] Broadbent, J., M. Palumbo and E. Woodman (2006), The Shift from Defined Benefit to Defined Contribution Pension Plans - Implications for Asset Allocation and Risk Management, https://www.bis.org/publ/wgpapers/cgfs27broadbent3.pdf.
[77] Caldara, D. and M. Iacoviello (2022), “Measuring Geopolitical Risk”, American Economic Review, Vol. 112/4, pp. 1194-1225, https://doi.org/10.1257/aer.20191823.
[81] Catalan, M., S. Fendoglu and T. Tsuruga (2024), “A gravity model of geopolitics and financial fragmentation”, IMF Working Papers 2024 196, https://doi.org/10.5089/9798400280337.001.
[79] Catalán, M. and T. Tsuruga (2023), Geopolitics and financial fragmentation: Implications for macro-financial stability, https://www.imf.org/-/media/files/publications/gfsr/2023/april/english/ch3.pdf.
[13] Çelik, S., G. Demirtaş and M. Isaksson (2020), Corporate Bond Market Trends, Emerging Risks and Monetary Policy, OECD Capital Market Series, http://www.oecd.org/corporate/Corporate-Bond-Market-Trends-Emerging-Risks-and-Monetary-Policy.htm.
[16] Chaudhary, M., J. Fu and H. Zhou (2024), “Anatomy of the Treasury Market: Who Moves Yields?”, Olin Business School Center for Finance & Accounting Research Paper, Vol. 2024/14, https://doi.org/10.2139/ssrn.5021055.
[69] Chen, Q. and J. Choi (2022), “Reaching for yield and the cross section of bond returns”, Management Science, Vol. 70/8, pp. 5226-5245, https://doi.org/10.1287/mnsc.2023.4920.
[59] Choi, J. and M. Kronlud (2017), “Reaching for yield in corporate bond mutual funds”, The Review of Financial Studies, Vol. 31/5, pp. 1930-1965, https://doi.org/10.1093/rfs/hhx132.
[84] Coibion, O. et al. (2025), Worrying about war: geopolitical risks weigh on consumer sentiment, https://www.ecb.europa.eu/press/blog/date/2025/html/ecb.blog20250407~7023432957.en.html.
[75] Constantini, M. and R. Sousa (2022), “What uncertainty does to euro area sovereign bond markets: Flight to safety and flight to quality”, Journal of International Money and Finance, Vol. 122, https://doi.org/10.1016/j.jimonfin.2021.102574.
[74] Converse, N. and E. Mallucci (2025), “Geopolitical Risk: When it Matters; Where it Matters. Evidence from International Portfolio Allocations”, https://www.esrb.europa.eu/news/schedule/2025/pdf/2025-06-30-awg-mpag8-annual-workshop-session-3-paper-nathan-converse-enrico-mallucci.pdf.
[61] Czech, R. and M. Robert-Sklar (2019), “Investor behaviour and reaching for yield: Evidence from the sterling corporate bond market”, Financial Markets, Inst. & Inst., Vol. 28/5, pp. 347-379, https://doi.org/10.1111/fmii.12122.
[47] D’Amico, S. and T. King (2013), “Flow and stock effects of large-scale treasury purchases: Evidence on the importance of local supply”, Journal of Financial Economics, Vol. 108/2, pp. 425-448, https://doi.org/10.1016/j.jfineco.2012.11.007.
[15] Domanski, D., H. Shin and V. Sushko (2015), “The hunt for duration: not waving but drowning?”, BIS Working Papers No. 519, https://www.bis.org/publ/work519.htm.
[29] Du, W., K. Forbes and M. Luzzetti (2024), “Quantitative Tightening Around the Globe: What Have We Learned?”, NBER Working Paper No. 32321, https://doi.org/10.3386/w32321.
[32] ECB (2021), The structural impact of the shift from defined benefits to defined contributions, ECB Economic Bulletin, Issue 5/2021, https://www.ecb.europa.eu/press/economic-bulletin/focus/2021/html/ecb.ebbox202105_08~5b846b2f5a.en.html.
[43] Ellul, A., C. Jotikasthira and C. Lundblad (2011), “Regulatory pressure and fire sales in the corporate bond market”, Journal of Financial Economics, Vol. 101/3, pp. 596-620, https://doi.org/10.1016/j.jfineco.2011.03.020.
[6] Epp, A. and J. Gao (2025), “The increasing role of hedge funds in Government of Canada bond auctions”, Bank of Canada Staff Analytical Note 2025-22, https://www.bankofcanada.ca/2025/10/staff-analytical-note-2025-22/.
[17] Eren, E., A. Schrimpf and F. Xia (2023), “The Demand for Government Debt”, BIS Working Papers No. 1105, https://doi.org/10.2139/ssrn.4466154.
[49] Eser, F. et al. (2019), “Tracing the Impact of the ECB’s Asset Purchase Programme on the Yield Curve”, ECB Working Paper No. 2293, https://doi.org/10.2139/ssrn.3417070.
[41] European Commission (2017), Analysis of European Bond Markets, https://finance.ec.europa.eu/system/files/2017-11/171120-corporate-bonds-analytical-report_en.pdf.
[37] FDIC (2025), Board of Directors Meeting on November 25, 2025, https://www.youtube.com/watch?v=fEWoCW8PakI&list=PLKjunAD4uy3hF-vILs-DttyLmOgyiPx7V&index=4.
[25] Fed (2025), Minutes of the Federal Open Market Committee, March 18-19, 2025, https://www.federalreserve.gov/monetarypolicy/files/fomcminutes20250319.pdf.
[22] Fed (2017), Section 13. Powers of Federal Reserve Banks, https://www.federalreserve.gov/aboutthefed/section13.htm.
[21] Fed (2017), Section 14. Open-Market Operations, https://www.federalreserve.gov/aboutthefed/section14.htm?.
[78] Feng, C. et al. (2023), “Geopolitical risk and the dynamics of international capital flows”, Journal of International Financial Markets, Institutions and Money, Vol. 82, https://doi.org/10.1016/j.intfin.2022.101693.
[56] Financial Times (2025), How the Treasury market got hooked on hedge fund leverage, https://www.ft.com/content/0bf5bcc2-6ff1-4309-afbf-f470250a4721.
[95] Foxall, S. and L. Policino (2025), “Sovereign retail debt programmes and instruments: A review of country practices”, OECD Working Papers on Sovereign Borrowing and Public Debt Management, No. 10, OECD Publishing, Paris, https://doi.org/10.1787/e2a782d0-en.
[48] Froemel, M., M. Joyce and I. Kaminska (2022), The Local Supply Channel of QE: Evidence from the Bank of England’s Gilt Purchases, Bank of England Working Paper No. 980, https://doi.org/10.2139/ssrn.4130492.
[36] FSB (2024), Global Monitoring Report on Non-Bank Financial Intermediation, https://www.fsb.org/uploads/P161224.pdf.
[93] Giese, J. et al. (2023), “Preferred habitat investors in the UK government bond market”, Staff Working Paper No. 939, https://www.bankofengland.co.uk/-/media/boe/files/working-paper/2021/preferred-habitat-investors-in-the-uk-government-bond-market.pdf.
[3] Giglio, S. (ed.) (2025), “In Safe Hands: The Financial and Real Impact of Investor Composition over the Credit Cycle”, The Review of Financial Studies, Vol. 38/8, pp. 2275-2325, https://doi.org/10.1093/rfs/hhaf017.
[63] Gomes, F. et al. (2023), “Reaching for yield: evidence from households”, Discussion Paper No. 887, https://www.fmg.ac.uk/sites/default/files/2023-10/DP887.pdf.
[57] Guttmann, R. (2012), Central Banking in a Systemic Crisis: The Federal Reserve’s ‘Credit Easing’, Edward Elgar Publishing, https://doi.org/10.4337/9781849807364.00017.
[60] Hanson, S. and J. Stein (2015), “Monetary policy and long-term real rates”, Journal of Financial Economics, Vol. 115, pp. 429-448, https://doi.org/10.1016/j.jfineco.2014.11.001.
[62] Huang, J. et al. (2025), “Do investors reach for yield? Evidence from corporate bond mutual fund flows”, Journal of Empirical Finance, Vol. 83, https://doi.org/10.1016/j.jempfin.2025.101625.
[83] IMF (2023), Global Financial Stability Report, April 2023: Safeguarding Financial Stability amid High Inflation and Geopolitical Risks, https://www.imf.org/en/publications/gfsr/issues/2023/04/11/global-financial-stability-report-april-2023.
[90] IMF (2004), Guidelines for foreign exchange reserve management, https://www.imf.org/external/pubs/ft/ferm/guidelines/2004/081604.pdf.
[97] IRS (2025), Defined benefit plan, https://www.irs.gov/retirement-plans/defined-benefit-plan.
[18] Jansen, K. (2025), “Long-Term Investors, Demand Shifts, and Yields”, The Review of Financial Studies, Vol. 38/1, pp. 114-157, https://doi.org/10.1093/rfs/hhae071.
[19] Jansen, K., W. Li and L. Schmid (2024), “Granular Treasury Demand with Arbitrageurs”, NBER Working Paper No. 33243, https://doi.org/10.3386/w33243.
[23] Japan Ministry of Finance (2025), Debt Management Report 2025, https://www.mof.go.jp/english/policy/jgbs/publication/debt_management_report/2025/esaimu2025.pdf.
[28] Jiang, Z. and J. Sun (2024), “Quantitative Tightening with Slow-Moving Capital”, NBER Working Paper No. 32757, https://doi.org/10.2139/ssrn.4558016.
[58] Kashyap, A. et al. (2025), “Treasury market dysfunction and the role of the central bank”, Brookings Papers on Economic Activity 1, https://www.brookings.edu/wp-content/uploads/2025/03/BPEA-SP25_WEB_Kashyap_Stein_Wallen_Younger.pdf.
[86] Khoo, J. and A. Cheung (2021), “Does geopolitical uncertainty affect corporate financing? Evidence from MIDAS regression”, Global Finance Journal, Vol. 47, https://doi.org/10.1016/j.gfj.2020.100519.
[2] Koijen, R. (ed.) (2025), “Who Holds Sovereign Debt and Why It Matters”, The Review of Financial Studies, Vol. 38/8, pp. 2326-2361, https://doi.org/10.1093/rfs/hhaf031.
[20] Koijen, R. et al. (2021), “Inspecting the mechanism of quantitative easing in the euro area”, Journal of Financial Economics, Vol. 140/1, https://doi.org/10.1016/j.jfineco.2020.11.006.
[94] Koijen, R. and M. Yogo (2023), “Understanding the Ownership Structure of Corporate Bonds”, American Economic Review: Insights, Vol. 5/1, pp. 73-92, https://doi.org/10.1257/aeri.20210550.
[14] Koijen, R. and M. Yogo (2019), “A Demand System Approach to Asset Pricing”, Journal of Political Economy, Vol. 127/4, https://doi.org/10.2139/ssrn.2537559.
[50] Krishnamurthy, A. and A. Vissing-Jorgensen (2011), “The Effects of Quantitative Easing on Interest Rates: Channels and Implications for Policy”, NBER Working Paper No. 17555, https://doi.org/10.3386/w17555.
[68] Lian, C., Y. Ma and C. Wang (2018), “Low interest rates and risk taking: evidence from individual investment decisions”, https://doi.org/10.1093/rfs/hhy111.
[64] Maggio, M. and M. Kacperczyk (2017), “The unintended consequences of the zero lower bound policy”, Journal of Financial Economics, Vol. 123/1, pp. 59-80, https://doi.org/10.1016/j.jfineco.2016.09.006.
[35] Maureen, O. and X. Zhou (2025), “US Corporate Bond Markets: Bigger and (Maybe) Better?”, Vol. 39/2, pp. 215-234, https://doi.org/10.1257/jep.20251439.
[34] Morgan Stanley (2018), Basel III: Impact on the Money Markets, https://www.morganstanley.com/im/publication/insights/regulatory/regulatory_baseliiiimpactonthemm_en.pdf.
[44] Murray, S. and S. Nikolova (2021), “The Bond Pricing Implications of Rating-Based Capital Requirements”, Vol. 57/6, pp. 2177-2207., https://doi.org/10.1017/S0022109021000132.
[42] NAIC (2021), Capital Adequacy (E) Task Force, https://content.naic.org/sites/default/files/inline-files/2021-09-H%20Proposal.pdf.
[1] OECD (2025), Global Debt Report 2025: Financing Growth in a Challenging Debt Market Environment, OECD Publishing, Paris, https://doi.org/10.1787/8ee42b13-en.
[9] OECD (2025), Pension Markets in Focus 2025, OECD Publishing, Paris, https://doi.org/10.1787/b095d0a0-en.
[30] OECD (2024), Global Debt Report 2024: Bond Markets in a High-Debt Environment, OECD Publishing, Paris, https://doi.org/10.1787/91844ea2-en.
[8] OECD (2024), Pension Markets in Focus 2024, OECD Publishing, Paris, https://doi.org/10.1787/b11473d3-en.
[98] OECD (2023), Pensions at a Glance 2023: OECD and G20 Indicators, OECD Publishing, Paris, https://doi.org/10.1787/678055dd-en.
[88] OECD (2021), OECD Sovereign Borrowing Outlook 2021, OECD Publishing, Paris, https://doi.org/10.1787/48828791-en.
[92] Palacios, M. and M. Habib (2024), “Sovereign wealth funds and the euro area: preliminary evidence”, https://www.ecb.europa.eu/press/other-publications/ire/focus/html/ecb.irebox202406_01~e8b7c42b87.en.html.
[76] Papavassiliou, V. (2025), “On the relationship between geopolitical risks and euro area sovereign bond yields”, Finance Research Letters, Vol. 75, https://doi.org/10.1016/j.frl.2025.106877.
[24] Paret, A. and A. Weber (2019), “German Bond Yields and Debt Supply: Is There a “Bund Premium”?”, Working Paper No. 2019/235, https://doi.org/10.5089/9781513518329.001.
[27] Pelizzon, L., M. Subrahmanyam and D. Tomio (2025), “Central Bank–Driven Mispricing”, Journal of Financial Economics, Vol. 166, https://doi.org/10.1016/j.jfineco.2025.104004.
[4] Philippon, T. (2006), “The Bond Market’s q”, NBER Working Paper No. 12462, https://doi.org/10.3386/w12462.
[53] Pinter, G. (2023), “An anatomy of the 2022 gilt market crisis”, Bank of England Staff Working Paper No. 1019, https://doi.org/10.2139/ssrn.4407740.
[7] Rousova, L. and M. Giuzio (2019), Insurers’ investment strategies, https://www.ecb.europa.eu/pub/pdf/scpwps/ecb.wp2299~1d060f6979.en.pdf.
[72] Schaaf, J. (2025), “From hype to hazard: what stablecoins mean for Europe”, The ECB Blog, https://www.ecb.europa.eu/press/blog/date/2025/html/ecb.blog20250728~e6cb3cf8b5.en.html.
[39] Stefan Goghie, A. (2025), Hedge funds and government bond markets: Why is there a love story?, https://goghieas.substack.com/p/hedge-funds-and-government-bond-markets.
[73] US Treasury (2025), Treasury Presentation to TBAC, https://home.treasury.gov/system/files/221/CombinedChargesforArchivesQ22025.pdf.
[31] Wright, J. (2022), “The Extent and Consequences of Federal Reserve Balance Sheet Shrinkage”, Brookings Papers on Economic Activity, Vol. 53/2, pp. 259-275, https://www.brookings.edu/wp-content/uploads/2022/09/BPEA-FA22_WEB_Wright.pdf.
Data on the investor base of sovereign and corporate bonds are classified as shown in the Table below. Following each source’ methodology, all instruments are valued in market prices except for the Japanese Treasury discount bills; and, for the US, (a) the holdings of sovereign and corporate bonds of MMFs, ABS, and other financial business and (b) the holdings of corporate bonds of the monetary authority, and the federal government. This implies that some changes in aggregate holdings may be due to valuation effects.
Note: * this is not available from QSA, so it is supplemented by data from the ECB Balance Sheet Items datasets on Euro area MMF holdings of Euro area government and corporate securities ** for government bond this is adjusted for estimated hedge funds holdings, as explained in this Annex.
Adjustment of US Households’ holdings of treasuries
As laid out in (Barth et al., 2025[87]), the household’s category of the US Financial Accounts erroneously contains a significant volume of US Treasuries which are held by hedge funds domiciled in the Cayman Islands. The data used in this chapter are adjusted accordingly, thereby decreasing the volume of bonds held by households and increasing the amount held by the foreign sector. For 2017-2024, the adjusted data from (Barth et al., 2025[87]) are used. For 2012-2016, the proxy suggested in Appendix A of (Barth et al., 2025[87]) was applied: the difference between total hedge fund US treasury holdings according to the Financial Accounts and domestic hedge fund US treasury holdings according to the Fed-Z1 release. For the remaining years, 2007-2011 and 2025, the adjustment is extrapolated by regressing the available data for 2012-2024 on the Financial Accounts household series and then using the fitted values.
Corporate document analysis
Mentions of different risk indicators in corporate disclosures refer to corporate filings and earnings call transcripts by companies included in the Russell 3 000 (US), Euro STOXX 600 (Europe) and NIKKEI 225 (Japan) indices that are included in the Bloomberg Document Search database. Document types include annual, semi‑annual and quarterly reports, 10‑K, 10‑Q, 8‑K, 6‑K, as well as earnings call transcripts, conference presentation calls, shareholder meeting calls and company presentations. The results are obtained through natural language processing built into the Bloomberg search function. Keywords shown in graphs are complemented with near or exact synonyms to expand results, such as “bond”, “credit” and “loan” for “debt” and “world political” and “geo political” for “geopolitical”.
This annex provides a detailed breakdown of the motives and strategies for the entities identified in the chapter that are active in sovereign and corporate bond markets.
Central banks buy bonds issued by governments and sometimes private entities in their domestic currency as part of their monetary policy toolkit, in particular when policy rates are near zero. These purchases are made to ease financial conditions by lowering longer-term yields and injecting liquidity; this is known as quantitative easing (QE) (OECD, 2021[88]). QE typically involves central banks announcing a pre‑determined amount of bonds they intend to purchase in an upcoming period. Another version of this is called yield curve control. This is where a central bank commits to buying whatever amount of bonds the market wants to supply to achieve a certain yield level (Belz and Wessel, 2020[89]). Central banks may also purchase bonds as part of their mandates to maintain financial stability. These purchases are almost always made in the secondary market.
Purchases made as part of QE or yield curve control are typically defined as market-neutral, with purchases roughly proportional to the outstanding volumes across the curve, or also purchases in a particular part of the curve where central banks are targeting yield levels or making a financial stability intervention. When it comes to purchases of corporate bonds, there are typically credit rating restrictions attached.
As central banks are large, price‑agnostic buyers with theoretically unlimited balance sheets, the announcement or even speculation of bond purchases can put downward pressure on yields. As sovereign yields serve as the benchmark for the wider economy, this reduces borrowing costs also for households and businesses, an intended effect of QE.
QE programmes are designed to be temporary. They are unwound either by not reinvesting the proceeds of matured holdings or by actively selling part of their holdings in the secondary market, or a combination of the two, in a process known as quantitative tightening (QT). Central banks cease being net buyers when they stop new purchases, including the reinvestment of matured holdings, but only become net sellers when they actively sell part of their bond holdings.
Foreign central banks
Foreign central banks and, to a lesser extent, other authorities like central governments are portfolio managers of foreign exchange reserves. Foreign exchange reserves are assets denominated in foreign currencies.
Given that reserves are held for mostly precautionary purposes, central banks and governments typically manage reserve portfolios with the objectives of preserving capital and maintaining liquidity. Foreign reserve managers primarily seek to ensure that adequate foreign exchange reserves are available to meet a defined range of objectives (for example, maintaining a currency peg); and that liquidity, market and credit risks are controlled in a prudent manner. Subject to liquidity and other risk constraints, modest returns over the medium to long term on the funds invested may be targeted as a tertiary objective (IMF, 2004[90]). As a result, foreign exchange reserves are typically invested in highly liquid securities that are of high credit quality, like highly rated government bonds and T-bills, as well as securities issued by government agencies or sub-sovereigns (Aldridge, Sandhu and Tchamova, 2024[91]).
Sovereign wealth funds
Sovereign wealth funds (SWFs) are state‑owned investment funds or entities that are commonly established to manage governments’ foreign assets. They are typically categorised as stabilisation funds to finance budget deficits or balance‑of-payments needs, savings funds for future generations, pension reserve funds, or reserve investment corporations established to reduce the carrying costs of foreign exchange reserves (Palacios and Habib, 2024[92]). Like conventional asset managers, they buy and hold bonds to help stabilise their portfolios. Both sovereign and corporate bonds provide a predictable income stream that is also highly liquid, particularly where a fund holds assets denominated in reserve currencies. This income stream allows fund managers to make additional investments without using the fund’s capital.
SWFs are generally opaque in their objectives and strategies, and there is a paucity of data and limited transparency about their portfolios. However, net flows tend to be from poorer to richer countries. SWFs will often pivot into bonds to defend and diversify their portfolios in anticipation of a slowdown in equity markets. In particular, SWFs that have been established for macroeconomic stabilisation objectives are likely to be overweight in the most liquid bond markets.
Commercial banks
Commercial banks’ main role in bond markets is as intermediaries. They buy bonds in the primary market and sell them to investors in the secondary market. They also conduct market making, providing liquidity by quoting both buy (bid) and sell (ask) prices for a given bond, thus creating a market for other participants. They stand ready to buy and sell bonds almost instantly, ensuring that there is enough volume for trades to be executed efficiently. Commercial bank purchases should therefore not be directly interpreted as additional demand for bonds. Commercial banks have traditionally also played a key intermediatory role in corporate bond markets. However, this has decreased significantly since the 2008 financial crisis, partly as a result of more stringent regulatory requirements that made it more capital-intensive to hold corporate bonds, reducing intermediation and proprietary inventories.
Banks are therefore mostly buying on behalf of clients, which may include their own treasuries, who will use them like an asset manager, to balance their portfolio. Banks also hold safe and liquid government bonds as part of their capital as they are considered high-quality liquid assets with zero risk weighting. (OECD, 2025[1]). The higher supply of bonds and greater volumes of trading in the post-2022 period have also increased the use of banks’ balance sheets and reduced their regulatory headroom against the supplementary leverage ratio (SLR).
Insurance companies
Insurance companies are typically hold to maturity investors. They are generally heterogenous in terms of duration preference, with life insurers in particular favouring ultra long duration, whilst other insurance funds tend to have a natural habitat a little bit shorter on the curve (Giese et al., 2023[93]). In 2023, insurance companies, especially life insurers, were the largest institutional investors in the US corporate bond market (Koijen and Yogo, 2023[94]). Insurance funds invest in bonds for multiple reasons, which can be broadly split into three categories:
Liability matching: Insurers must pay out claims over time, often years or decades in the future, and bonds provide predictable income streams, making it easier to match assets with future liabilities.
Capital preservation: Insurers prioritise capital safety over high returns. Bonds, especially those issued by sovereigns and high-grade corporates, are typically less volatile than equities, helping preserve insurers’ ability to meet their obligations to policyholders.
Regulatory and rating constraints: Insurers are tightly regulated with capital adequacy frameworks (like Risk-Based Capital in the United States or Solvency II in the EU) making investment in lower-risk assets like bonds less capital-intensive. Meanwhile, credit rating agencies also tend to favour stable and conservative asset mixes for insurance companies.
Households and non-profit institutions serving households (NPISHs)
Households and NPISHs or “retail investors”, as they are often referred to, buy bonds for reasons including capital preservation and retirement planning (Foxall and Policino, 2025[95]). Bond holdings, notably by sovereign issuers, serve to ensure some returns on capital without too much risk exposure; retail investment in bond markets can therefore effectively be seen as a substitute for bank deposits and savings accounts offered by commercial banks. Patriotic sentiments can also be an important factor for retail investors in sovereign bond markets, especially in times of war or national disasters, or as a means for diaspora to invest in their home countries. Households tend to hold sovereign bonds to greater extent than corporate ones, as sovereign issuers often have dedicated programmes and/or products to target this investor group, whereas access to corporate bonds tends to be more restricted, with high minimum investment amounts.
Compared to other investor categories, retail investors are usually considered to be more price sensitive when deciding whether to buy, but more likely to hold to maturity when they do buy, whilst exhibiting greater home bias. Retail investors are more likely to buy when real rates are higher, as they are buying more for yield than liquidity. Where retail investors buy marketable products, they are less likely to trade relatively small fluctuations in the secondary market price and instead prefer to hold to maturity, where they receive a guaranteed return. Meanwhile, many sovereign retail debt products are non-marketable and therefore designed to be held to maturity, with sometimes fees or penalties for early redemption. Retail investors also typically have a stronger preference for investing in their local markets.
The recent resurgence of retail investors in sovereign bond markets has been driven by more attractive yields, greater market accessibility, and active efforts by issuers to diversify their investor bases amid heightened macrofinancial uncertainty. However, as an investor group, they may fail to raise significant levels of funding, particularly in jurisdictions where there are suitable and competitive alternative products available to them. For issuers with particularly high borrowing requirements, retail investors will likely only ever constitute a small share of the investor base (Boitreaud, Foxall and Policino, 2025[96]).
Pensions funds
Asset-backed pension funds, which have long-term investment horizons and predominantly clear payout schedules, are natural investors in bond markets. In general, they can be said to be less price‑sensitive than other investor groups and typically buy bonds to hold to maturity, favouring higher duration and, in certain jurisdictions, inflation-linked instruments. It is important, however, to distinguish between different types of funds, in particular between defined benefit (DB) and defined contribution (DC) funds, because their incentives and behaviours as investors, notably their allocations to fixed income, differ substantially. The global shift from DB to DC structures is therefore having profound implications for sovereign and corporate bond markets.
Defined benefit (DB) pensions
A DB pension scheme provides a fixed, pre‑established benefit for employees at retirement (IRS, 2025[97]). As such, they are typically required to structure their assets in a way that matches the duration of their liabilities. Consequently, they are significant holders of long-dated investment grade bonds, including inflation-linked bonds, as the market price of these assets respond to changes in interest rates and inflation in the same way as the scheme’s liabilities. Periodic coupon payments also generate predictable cash flows to cover the scheme’s cash outflows.
Defined contribution (DC) pensions
A DC pension scheme is a type of private pension that the employee and their employer contribute to on a regular basis. They define how much and when to pay into it. That pot of money is then invested over the life of the pension, with the employee carrying both asset value risk and longevity risk (OECD, 2023[98]).
Bonds are often used to help spread the risk in DC pensions, particularly as individuals get closer to retirement. As DC pension schemes are fully funded and focus on return maximisation and investment flexibility, rather than duration matching, DC assets are generally weighted towards higher-returning assets such as listed equities; they may also be more likely to invest in higher yielding corporate bonds.
Investment funds
Conventional asset managers manage investment funds on behalf of clients, including through mutual funds, exchange traded funds (ETFs) and private accounts. For present purposes, this category covers asset managers other than pension funds, insurers and hedge funds. These investors are characterised by higher price‑sensitivity than pension and insurance funds, but a greater tendency to buy to hold than more opportunistic investors like hedge funds.
Investing in bonds provides diversification benefits for asset managers. These benefits arise from the usually low correlations of bonds with other major asset classes, such as equities. Floating-rate and inflation-linked bonds can also be used to hedge inflation risk. Bonds have regular cash flows, which are beneficial for the purposes of funding future liabilities.
Hedge funds
Hedge funds usually buy and sell bonds to profit from changes or discrepancies in prices between similar assets. Unlike pension and insurance funds, they do not hold onto bonds for liquidity or yield. Instead, they are trying to generate positive alpha, using leverage, derivatives, and timing.
Hedge funds have generally responded to the higher volumes of bond issuance by buying more; this is particularly the case in large sovereign bond markets, such as Canada, Germany and the United States, because of their volume‑based business models. Investors other than hedge funds are generally more limited by their cash on hand and tend to use less leverage (Epp and Gao, 2025[6]).
Broadly speaking, the approach taken by hedge funds in sovereign bond markets is to identify a perceived mispricing and take a large position on that perceived mispricing being corrected, as observed in hedge funds’ substantial use of repo-based leverage. Hedge funds also often use relative value strategies, trading on the perceived mispricing between bonds and either:
Futures – known as the cash-futures basis trade
Swaps of a similar term – known as the asset-swap trade
Bonds of the same issuer – known as intra‑issuer relative‑value trades
Other sovereign bonds – known as the cross-curve trades
Hedge funds also commonly use strategies based on macroeconomic views (e.g. position on level or shape of the yield curve) or bond pricing at auction close. They also use government bonds as collateral for derivatives.
Because mispricing opportunities in liquid bond markets are typically relatively small in size, hedge funds have to take larger positions to achieve their desired return. In the cash-futures basis trade, for example, which involves a long position in a bond and a short position in a futures contract, hedge funds use leverage to amplify their profits as the spread between the two securities is generally relatively narrow.
Higher free float makes it easier for hedge funds to take larger long positions. This is because it increases the available bonds to hold for those positions and provides hedge funds with a larger base of assets to repo out to fund their positions. In addition, a larger known future supply event (i.e. bond auction or syndication) can facilitate larger short positions. That’s because the larger supply allows hedge funds to obtain the required bonds more easily in the repo market. Further, when the bonds are more liquid, the transaction costs for entering and exiting positions are lower.
← 1. The category “Other” includes ABS, broker/dealers, closed-end funds, credit unions, ETFs, GSE and agency, holding companies, mutual funds, other financial businesses, mortgage REITs, financial companies, non-financial corporate business, state and local government, federal government for US; collectively managed trusts, securities investment trusts, other financial intermediaries, financial auxiliaries, public captive financial institutions, domestic non-financial sector, general government for Japan; non-MMF investment funds, other financial intermediaries except insurance corporations and pension funds, financial auxiliaries, captive financial institutions and moneylenders, non-financial companies, general government for the euro‑area.
← 2. For instance, around one‑third of Euro area households’ bond holdings were held via investment funds between 2015 and 2025.
← 3. It should also be noted that due to national accounts classifications, certain de facto government pension funds are classified within the General Government sector. Accordingly, this report considers the Japanese Government Pension Investment Fund into the Pension Funds category, rather than into the General Government. No further modifications of this kind were made to the dataset.
← 4. According to the 2025 OECD Survey on Primary Markets, six OECD countries issued new retail products in the 12 months leading to October 2025 and six OECD countries plan to issue new retail products in the following12 months. Since 2020, Italy has launched several new products tailored to retail investors: BTP Futura in 2020, BTP Valore in 2023. The latter with several variants such as the one mentioned in the text, which was introduced in 2025.
← 5. Global Systemically Important Banks (G-SIBs) are banks whose failure could significantly disrupt the global financial system and are therefore subject to stricter regulatory requirements. G‑SIBs are identified by the Financial Stability Board (FSB) and published each year using a multi‑indicator scoring system. 2025 List of Global Systemically Important Banks (G-SIBs) - Financial Stability Board.
← 6. Scarcity effects, free‑float compression and crowding-out of foreign investors arise primarily in government bond markets and do not generalise to corporates.
← 7. The distinction between price‑sensitive and price‑insensitive investors here is of course simplified. In reality it is likely regime‑dependent since these traits can change with market and policy conditions.
← 8. Spread revenue refers to the difference between the actual earnings on investments made by the insurer and the crediting rate guaranteed to participants for that period. The crediting rate is subject to the minimum guaranteed rate stated in the insurance contract.