Keynote speech by Philip R. Lane, Member of the Executive Board of the ECB, at the Government Borrowers Forum 2025

Dublin, 11 June 2025

I am grateful for the invitation to contribute to the Government Borrowers Forum. I will use my time to cover three topics.^[1] First, I will briefly discuss last week’s monetary policy decision.^[2] Second, I will describe some current features of the euro area bond market.^[3] Third, I will outline some innovations that might expand the scope for euro-denominated bonds to serve as safe assets in global portfolios.

Monetary policy

At last week’s meeting, the Governing Council decided to lower the deposit facility rate (DFR) to two per cent. The baseline of the latest Eurosystem staff projections foresees inflation at 2.0 per cent in 2025, 1.6 per cent in 2026 and 2.0 per cent in 2027; output growth is foreseen at 0.9 per cent for 2025, 1.2 per cent in 2026 and 1.3 per cent in 2027. The lower inflation path in the June projections compared to the March projections reflects the significant movements in energy prices and the exchange rate in recent months. These relative price movements both have a direct impact on inflation but also an indirect impact via the impact of lower input costs and a lower cost of living on the dynamics of core inflation and wage inflation.

The June projections were conditioned on a rate path that included a quarter-point reduction of the DFR in June: model-based optimal policy simulations and an array of monetary policy feedback rules indicated a cut was appropriate under the baseline and also constituted a robust decision, remaining appropriate across a range of alternative future paths for inflation and the economy. By supporting the pricing pressure needed to generate target-consistent inflation in the medium-term, this cut helps ensure that the projected negative inflation deviation over the next eighteen months remains temporary and does not convert into a longer-term deviation of inflation from the target. This cut also guards against any uncertainty about our reaction function by demonstrating that we are determined to make sure that inflation returns to target in the medium term. This helps to underpin inflation expectations and avoid an unwarranted tightening in financial conditions.

The robustness of the decision is also indicated by a set of model-based optimal policy simulations conducted on various combinations of the scenarios discussed in the Eurosystem staff projections report, even when also factoring in upside scenarios for fiscal expenditure. A cut is also indicated by a broad range of monetary policy feedback rules. By contrast, leaving the DFR on hold at 2.25 per cent could have triggered an adverse repricing of the forward curve and a revision in inflation expectations that would risk generating a more pronounced and longer-lasting undershoot of the inflation target. In turn, if this risk materialised, a stronger monetary reaction would ultimately be required.

Especially under current conditions of high uncertainty, it is essential to remain data dependent and take a meeting-by-meeting approach in making monetary policy decisions. Accordingly, the Governing Council does not pre-commit to any particular future rate path.

The euro area bond market

Let me now turn to a longer-run perspective by inspecting developments in the bond market. In the first two decades of the euro, nominal long-term interest rates in the euro area were, by and large, on a declining trend from the start of the currency bloc until the outbreak of the pandemic (Chart 1). The ten-year overnight index swap (OIS) rate, considered as the ten-year risk-free rate in the euro area, declined from 6 percent in early 2000 to -50 basis points in 2020, a trend matched by the 10-year GDP-weighted sovereign bond yield.^[4] The economic recovery from the pandemic and the soaring energy prices in response to the Russian invasion in Ukraine caused surges in inflation which led to an increase of interest rates. The recent stability of these long-term rates suggests that markets have seen the euro area economy gradually moving towards a new long-term equilibrium following the peak of annual headline inflation in October 2022, as past shocks have faded.

A term structure model makes it possible to decompose OIS rates into a term premium component and an expectations component. For the ten-year OIS rate, the expectations component reflects the expected average ECB policy rate over the next ten years and is affected by ECB’s policy decisions on interest rates and communication about the future policy path (e.g., in the form of explicit or implicit forward guidance). The term premium is a measure of the estimated compensation investors demand for being exposed to interest rate risk: the risk that the realised policy rate can be different from the expected rate.

The decline of long-term rates in the first two decades of the euro and the rapid increase in 2022 were driven by both the expectations component and the term premium (Charts 2 and 3). The premium was estimated to be largely positive in the early 2000s, understood as a sign that the euro area economy was mostly confronted with supply-side shocks. Starting with the European sovereign debt crisis, the euro area was more and more characterised as a demand-shock dominated economy, in which nominal bonds act as a hedge against future crises and thus investors started requiring a lower or even negative term premium as compensation to hold these assets.^[7] The large-scale asset purchases of the ECB under the APP reinforced the downward pressure on the term premium. By buying sovereign bonds (and other assets), the ECB reduced the overall amount of duration risk that had to be borne by private investors, reducing the compensation for risk.^[8] With demand and supply shocks becoming more balanced again and central banks around the world normalising their balance sheet holdings of sovereign bonds in recent years, the term premium estimate turned positive again in early 2022 and continued to inch up through the first half of 2023. As it became clear in the second half of 2023 that upside risk scenarios for inflation were less likely, the term premium fell back to some extent and has been fairly stable since.

Different to the ten-year maturity, very long-term sovereign spreads did not experience the same pronounced negative trend. From the inception of the euro until 2014, the thirty-year euro area GDP-weighted sovereign yield fluctuated around 3 percent. The decline to levels below 2 percent after 2014 and around 0.5 percent in 2020 reflect declining nominal risk-free rates more generally but also coincide with the announcements of large-scale asset purchases (PSPP and PEPP). Likewise, the upward shift back to above 3 percent during 2022 occurred on the back of rising policy rates and normalising central bank balance sheets.

In the run-up to the global financial crisis, sovereign yields in the euro area were very much aligned between countries and also with risk-free rates (Chart 4). With the onset of the global financial crisis and later the European sovereign debt crisis, sovereign spreads for more vulnerable countries soared as investors started to discriminate between euro area countries according to their perceived creditworthiness.

On top of the efforts of European sovereigns to consolidate their public finances, President Draghi’s 2012 “whatever it takes” speech and the subsequent announcement of Outright Monetary Transaction (OMTs) marked a turning point in the euro area sovereign debt crisis. Sovereign spreads came down from their peaks but have kept some variation across countries ever since.

The large-scale asset purchases under the APP and PEPP further compressed sovereign spreads. During the pandemic and the subsequent monetary policy tightening, the flexibility in PEPP and the creation of the Transmission Protection Instrument (TPI) supported avoiding fragmentation risks in sovereign bond markets. The extraordinary demand for sovereign bonds as collateral at the beginning of the hiking cycle, at a time when central bank holdings of these bonds were still high, resulted in the yields of German bonds, which are the most-preferred assets when it comes to collateral, declining far below the risk-free OIS rate in the course of 2022. These tensions eased as collateral scarcity reversed.^[9]

This year, bond yields and bond spreads in the euro area have been relatively stable, despite significant movements in some other bond markets. This can be interpreted as reflecting a balancing between two opposing forces: in essence, the typical positive spillover across bond markets has been offset by an international portfolio preference shift towards the euro and euro-denominated securities. This international portfolio preference shift is likely not uniform and is some mix of a pull back by European investors towards the domestic market and some rebalancing by global investors away from the dollar and towards the euro. More deeply, the stability of the euro bond market reflects a high conviction that euro area inflation is strongly anchored at the two per cent target and that the euro area business cycle should be relatively stable, such that the likely scale of cyclical interest rate movements is contained. It also reflects growing confidence that the scope for the materialisation of national or area-wide fiscal risks is quite contained, in view of the shared commitment to fiscal stability among the member countries and the demonstrated capacity to react jointly to fiscal tail events.^[10]

In understanding the dynamics of the bond market, it is also useful to examine the distribution of bond holdings across sectors. The largest euro-area holder sectors are banks, insurance corporations and pension funds (ICPF) and investment funds, while non-euro area foreign investors also are significant holders (Chart 5). The relative importance of the sectors differs between countries. Domestic banks and insurance corporations play a relatively larger role in countries like Italy and Spain, while non-euro area international investors hold relatively larger shares of debt issued by France or Germany.

Since the start of the APP in early 2015, the Eurosystem increased its market share in euro area sovereign bonds from about 5 per cent of total outstanding debt to a peak of 33 per cent in late 2022. Net asset purchases by the Eurosystem were stopped in July 2022, while the full reinvestment of redemptions ceased at the end of that year: by Q1 2025, the Eurosystem share had declined to 25 per cent. The increase in Eurosystem holdings during the QE period was mirrored by falling holdings of banks and non-euro area foreign investors. The holding share of banks declined from 22 per cent in 2014 to 14 per cent at the end of 2022, while the share held by foreign investors fell from 35 per cent to 25 per cent over the same period.

ICPFs have consistently held a significant share of the outstanding debt, especially at the long-end of the yield curve. Since 2022, following the end of full reinvestments under the APP, more price-sensitive sectors, such as banks, investment funds and private foreign investors, have regained some market share. Holdings by households have also shown some noticeable growth in sovereign bond holdings, driven primarily by Italian households.^[11] In summary, the holdings statistics show that the bond market has smoothly adjusted to the end of quantitative easing. In particular, the rise in bond yields in 2022 was sufficient to attract a wide range of domestic and global investors to expand their holdings of euro-denominated bonds.^[12]

To gain further insight into the recent dynamics of the euro area bond market, it is helpful to look at recent portfolio flow data and bond issuance data. Market data on portfolio flows^[13] highlights a repatriation of investment funds in bonds by domestic investors during March, April, and May, contrasting sharply with 2024 trends, while foreign fund inflows into euro area bonds during the same period surpassed the 2024 average (Chart 6). Simultaneously, EUR-denominated bond issuance by non-euro area corporations has surged in 2025, reaching nearly EUR 100 billion year-to-date compared to an average of EUR 32 billion over the same period in the past five years (Chart 7).

Expanding the pool of safe assets

These developments (stable bond yields, increased foreign holdings of euro-denominated bonds) have naturally led to renewed interest in the international role of the euro.^[14]

The euro ranks as the second largest reserve currency after the dollar. However, the current design of the euro area financial architecture results in an under-supply of the safe assets that play a special role in investor portfolios.^[15] In particular, a safe asset should rise in relative value during stress episodes, thereby providing essential hedging services.

Since the bund is the highest-rated large-country national bond in the euro area, it serves as the main de facto safe asset but the stock of bunds is too small relative to the size of the euro area or the global financial system to satiate the demand for euro-denominated safe assets. Especially in the context of much smaller and less volatile spreads (as shown in Chart 4), other national bonds also directionally contribute to the stock of safe assets. However, the remaining scope for relative price movements across these bonds means that the overall stock of national bonds does not sufficiently provide safe asset services.

In principle, common bonds backed by the combined fiscal capacity of the EU member states are capable of providing safe-asset services. However, the current stock of such bonds is simply too small to foster the necessary liquidity and risk management services (derivative markets; repo markets) that are part and parcel of serving as a safe asset.^[16]

There are several ways to expand the stock of common bonds. Just as the Next Generation EU (NGEU) programme was financed by the issuance of common bonds jointly backed by the member states, the member countries could decide to finance investment European-wide public goods through more common debt.^[17] From a public finance perspective, it is natural to match European-wide public goods with common debt, in order to align the financing with the area-wide benefits of such public goods. If a multi-year investment programme were announced, the global investor community would recognise that the stock of euro common bonds would climb incrementally over time.

In addition, in order to meet more quickly and more decisively the rising global demand for euro-denominated safe assets, there are a number of options in generating a larger stock of safe assets from the current stock of national bonds. Recently, Olivier Blanchard and Ángel Ubide have proposed that the “blue bond/red bond” reform be re-examined.^[18] Under this approach, each member country would ring fence a dedicated revenue stream (say a certain amount of indirect tax revenues) that could be used to service commonly-issued bonds. In turn, the proceeds of issuing blue bonds would be deployed to purchase a given amount of the national bonds of each participating member state. This mechanism would result in a larger stock of common bonds (blue bonds) and a lower stock of national bonds (red bonds).

While this type of financial reform was originally proposed during the euro area sovereign debt crisis, the conditions today are far more favourable, especially if the scale of blue bond issuance were to be calibrated in a prudent manner in order to mitigate some of the identified concerns. In particular, the euro area financial architecture is now far more resilient, thanks to the significant institutional reforms that were introduced in the wake of the euro area crisis and the demonstrated track record of financial stability that has characterised Europe over the last decade. The list of reforms include: an increase in the capitalisation of the European banking system; the joint supervision of the banking system through the Single Supervisory Mechanism; the adoption of a comprehensive set of macroprudential measures at national and European levels; the implementation of the Single Resolution Mechanism; the narrowing of fiscal, financial and external imbalances; the fiscal backstops provided by the European Stability Mechanism; the common solidarity shown during the pandemic through the innovative NGEU programme; the demonstrated track record of the ECB in supplying liquidity in the event of market stress; and the expansion of the ECB policy toolkit (TPI, OMT) to address a range of liquidity tail risks. ^[19] In the context of the sovereign bond market, these reforms have contributed to less volatile and less dispersed bond returns.

As emphasised in the Blanchard-Ubide proposal, there is an inherent trade off in the issuance of blue bonds. In one direction, a larger stock of blue bonds boosts liquidity and, if a critical mass is attained, also would trigger the fixed-cost investments need to build out ancillary financial products such as derivatives and repos. In the other direction, too-large a stock of blue bonds would require the ringfencing of national tax revenues at a scale that would be excessive in the context of the current European political configuration in which fiscal resources and political decision-making primarily remains at the national level. As emphasised in the Blanchard-Ubide proposal, this trade-off is best navigated by calibrating the stock of blue bonds at an appropriate level.

In particular, the Blanchard-Ubide proposal gives the example of a stock of blue bonds corresponding to 25 per cent of GDP. Just to illustrate the scale of the required fiscal resources to back this level of issuance: if bond yields were on average in the range of two to four per cent, the servicing of blue bond debt would require ringfenced tax revenues in the range of a half per cent to one per cent of GDP. While this would constitute a significant shift in the current allocation of tax revenues between national and EU levels, this would still leave tax revenues predominantly at the national level (the ratio of tax revenues to GDP in the euro area ranges from around 20 to 40 per cent). The shared payoff would be the reduction in debt servicing costs generated by the safe asset services provided by an expanded stock of common debt.

An alternative, possibly complementary, approach that could also deliver a larger stock of safe assets from the pool of national bonds is provided by the sovereign bond backed securities (SBBS) proposal.^[20] The SBBS proposal envisages that financial intermediaries (whether public or private) could bundle a portfolio of national bonds and issue tranched securities, with the senior slice constituting a highly-safe asset. The SBBS proposal has been extensively studied (I chaired a 2017 ESRB report) and draft enabling legislation has been prepared by the European Commission.^[21] Just as with the blue/red bond proposal, sufficient issuance scale would be needed in order to foster the market liquidity needed for the senior bonds to act as highly liquid safe assets.

In summary, such structural changes in the design of the euro area bond market would foster stronger global demand for euro-denominated safe assets. A comprehensive strategy to expand the international role of the euro and underpin a European savings and investment union should include making progress on this front.