The ECB could slow down QT in response to currently tight financial conditions, but should communicate its intentions with care.

Since early 2023, the European Central Bank has substantially reduced the size of its balance sheet, without any acute stress in euro-area bond markets. The balance sheet had ballooned as a consequence of unconventional monetary policy measures implemented by the ECB from 2015 to address inflation persistently below its 2 percent target, and then to respond to the COVID-19 pandemic shock. The bank has also instituted changes to its operational framework as part of a transition from an era of quantitative easing (QE) – asset purchases – to quantitative tightening (QT), or the releasing of those assets back into the market.

The ECB is doing this when the stretched public finances of European Union countries are again coming under pressure because of higher borrowing costs induced by the energy and commodity price shock emanating from the conflict in the Middle East.

Should the ECB slow the reduction of its holdings of sovereign bonds? The main justification for slowing would be to offset the unintended tightening of financial conditions that has already taken place, while also reducing the risk of being forced to ease monetary policy more aggressively through direct intervention, either by resorting to its yet untested Transmission Protection Instrument (TPI), or a more conventional asset purchase programme if the increase in government borrowing costs is not confined to a few countries.

ECB balance-sheet unwinding is going well

The ECB began reducing its holdings of government bonds in March 2023 by not fully reinvesting the proceeds from maturing bonds (‘passive unwinding’). At its end-2022 peak, it held over €5 trillion in government bonds, holdings that are intended to directly impact financial conditions through their impact on government bond yields. This impact has been substantial: together with negative policy rates, yields on long maturity bonds averaged less than 1 percent across the euro area for an extended period.

At its peak, the ECB’s aggregate asset portfolio – including special liquidity provision and other refinancing operations adding another €2 trillion to its balance sheet – amounted to 50 percent of euro-area GDP. This was a much larger proportion than the balance sheets of the US Federal Reserve (30 percent of GDP) or the Bank of England (40 percent). The Fed and Bank of England balance sheets mainly comprised their asset purchases. The ECB’s liquidity provision facilities were large, but for shorter maturities and rolled automatically, without much market impact, mainly because the banking system had adequate reserves supplied through ECB bond purchases. 

The relatively shorter average maturity (duration) of the ECB’s bond holdings implies that it can reduce its balance sheet faster than, for example, the Bank of England. It can do this passively by not reinvesting proceeds from maturing bonds. The average maturity of the ECB portfolio was just over seven years before it began unwinding. Therefore, from a start in 2023, the ECB should halve its bond portfolio passively by 2030. The ECB could reduce the pace of QT if markets are not able to absorb more debt without a material rise in yields.

But would less QT be consistent with the ECB’s current operational framework for monetary policy normalisation – essentially how it will move, over time, to a smaller balance sheet?

Normalisation of monetary policy: the ECB’s operational framework

In March 2024, the ECB’s Governing Council (GC) laid out the contours of a new framework to implement monetary policy as ‘excess liquidity’ continues to decline gradually, and its balance sheet shrinks. Importantly, none of the changes introduced by the new framework precluded changing the pace of QT, or indeed resuming QE, if appropriate. The ECB committed to retaining the flexibility to use all its instruments, including longer-term refinancing operations (providing liquidity), asset purchases, negative interest rates and forward guidance.

Maintaining full flexibility will be helpful for the ECB if the external shock of the crisis over Iran impairs the even transmission of monetary policy across the monetary union, as well as raising yields.

The March 2024 revisions to the framework included two important changes, which are still to be introduced:

1. The ECB will move away from providing liquidity on demand – when the banking system was supplied with ample reserves during QE – to a system in which it will charge a higher interest rate for banks that need to borrow. This will incentivise banks to more actively manage their liquidity, including transacting with other banks.
2. The size and composition of assets the ECB will hold when, in principle, QT has run its course – what the ECB terms its longer-term Structural Portfolio. Importantly, this will be larger than before QE, because banks will want to hold larger buffers because of regulatory and technological changes and higher precautionary buffers. It will likely be substantially more than the pre-QE level of €800 billion but less than the current level of over €3.5 trillion.

The ECB’s future bond portfolio will, however, comprise much shorter maturities, largely because its purpose would be to meet the banking system’s short-term reserve needs. Moreover, the ECB wants to avoid risks of large losses on long-term bonds when interest rates change, such as happened when interest rates increased substantially following the COVID-19 and Russia-Ukraine shocks (Gross and Shamsfakhr, 2020).

Would slower QT lead to lower yields?

It is empirically difficult to isolate the impact on market yields of reductions in central bank bond holdings from other factors. The primary impact is through changes in the composition of investors’ portfolios (the portfolio balance effect): as a central bank buys bonds, the composition of investors’ portfolios change, inducing investors to buy other assets. A second impact is through expectations of future changes in policy rates that asset purchases might imply (the signalling effect). Research on the short-term impact of asset-purchase announcements on yields often ignores the possible longer-term (cumulative) stock effect as the central bank implements the programme.

Broader evidence on the impact of QT on yields, encompassing both the announcement and cumulative impact, shows QT has a sizeable impact on yields. Wenxin et al (2024) examined the early impact of QT in Australia, Canada, the euro area, New Zealand, Sweden, the UK and the US, finding sizeable announcement impacts (partly because all these central banks were also raising interest rates), but also strong cumulative effects. Though markets functioned in an orderly way and absorbed the additional supply, Wenxin et al (2024) cautioned that this resilience may not hold up in the face of new shocks and continuing reductions in central-bank balance sheets. 

Empirical research on the ECB suggests QE has a bigger negative impact on yields than the positive impact of QT. Altavilla et al (2021) estimated that a 10 percent GDP asset purchase programme could reduce yields by 65 basis points. Akkaya et al (2024) suggested that a €1 trillion reduction in its balance sheet (10 percent of euro-area GDP) would lead to a 35 basis point increase in German 10-year yields with a larger increase in Italian (50 basis points) and Spanish (45 basis points) 10-year yields. 

Part of this difference depends on which bonds the markets give up during QE – typically spanning the entire yield curve. By contrast, when a central bank stops reinvesting proceeds from maturing bonds, it is essentially acting only at the short end of the yield curve (maturing bonds expire). The impact on yields for longer maturities then depends on how governments refinance maturing debt, which will vary across issuers, and on aggregate debt issuance. 

In the current environment of rising energy prices and some governments shielding households from part of the shock, it is more likely than not that debt issuance will rise.

How much debt fiscal authorities’ issue will arguably have a bigger impact on yields than suggested by estimates based on how much yields have risen since early 2023 because of QT. Higher borrowing costs now, compared to previous periods, have increased debt-service costs, implying that governments must run larger primary surpluses.

The already underway larger reduction in the ECB’s balance sheet could have a sizeable impact. Schnabel (2025) estimated that from March 2023 to October 2025, QT increased yields by a GDP-weighted average of 60 basis points across Germany, France, Italy and Spain. These estimates also imply that yields before the start of the Ukraine war were about 20 basis points below what they would have been without QE since 2015, and that markets will demand higher yields if QT continues at the current pace.

Why the ECB should consider more flexible QT

Any tightening induced by a shock – such as the 40-50 basis point increase in yields since the US/Iran conflict started – implies a more restrictive monetary stance than intended by policymakers. This is why central banks, including the ECB, should not (and indeed do not) commit to a pre-determined schedule of balance-sheet reduction independent of financial conditions.

But with inflation projected to remain above the ECB’s 2 percent price-stability objective, any policy that indicates a more accommodative stance would raise questions about credibility. Any easing of policy rates in the current environment might increase longer-term yields if it resulted in higher expected inflation. Consequently, the ECB would not reduce policy rates – and markets would not expect it to.

Less QT, however, would arguably not trigger the same concerns. Markets can typically absorb more government debt held by central banks when they are not asked to absorb more new debt issuance from governments. With the Iran-related energy price shock, governments are likely to issue more debt, and deficits will rise over time because of higher debt-service costs.

The case for a short-term QT pause is stronger now because the energy shock will add to already high debt burdens – four euro-area countries (Belgium, France, Italy and Spain) have debt/GDP ratios above 100 percent. Unlike previous periods of rising borrowing costs, yields this time are also rising for Belgium, France and Germany. Some countries are already introducing tax cuts to reduce energy bills; though small at this stage, these measures are not targeted (Bruegel Dataset, 2026). Despite an outline agreement to end the US/Iran conflict, energy prices are unlikely to return to pre-war levels for at least a few months because of damage to production facilities, transportation bottlenecks and supply-chain disruptions.

Slowing down QT while raising policy interest rates would not be unprecedented. Rates were raised in the second half of 2022 before the start of QT, and reinvestment from maturing bonds under the ECB’s pandemic emergency purchase programme (PEPP) continued until early 2023. In other words, the ECB has raised rates before to address inflation concerns without reducing its balance sheet.

Risk of fiscal dominance

Nevertheless, a potential concern, including for the ECB, is that slowing QT could exacerbate fiscal dominance – the notion that monetary policy actions would be determined not only by price-stability objectives, but would be accommodating governments’ financing needs – notwithstanding that slower QT would be a short-term measure.

Limiting the rise in borrowing costs below levels that the market would determine under the current QT path would undoubtedly help governments and increase the risk of undermining the ECB’s credibility. A temporary pause would arguably minimise this risk.

First, it would not alleviate the need for fiscal adjustment. It is designed to help with the additional pressures on borrowing costs from the energy shock. Governments would not obtain any relief from numerous other structural demands on their fiscal resources, such as higher defence expenditure commitments, net-zero goals, rising ageing-related health and social benefits and security. All these demands will entail fiscal adjustment through trade-offs: spending will have to be cut in some areas.

Second, it would be consistent with the ECB’s mandate because it is simply a change in the pace of balance-sheet unwinding, and not any significant change in the monetary policy stance, which, as the ECB frequently reiterates, is primarily determined by its policy rate. A QT slowdown would be a response to a portion of higher bond yields resulting from an increase in the term premium and additional issuance because of this shock; it would not offset the increase in yields arising from higher expected inflation.

But how does the ECB guard against losing credibility? How does it convince markets (and governments) that this would be a temporary pause? It could leave unspecified how long it plans to slow down QT, not target any yield levels and emphasise that it is simply adjusting to tighter financial conditions caused by an external shock. These arguments may not be fully persuasive because markets will assign some probability that the ECB will be inclined to (or persuaded through political pressure) repeat such accommodation as governments continue to face fiscal pressures. But nor is the alternative of not adjusting policy in the face of tighter financial conditions because of a risk of fiscal dominance.

Moreover, doing nothing could mean waiting for the adverse impacts of the energy shock to show up in a material growth slowdown, which in turn would revise the ECB’s inflation projections. If policymakers are persuaded by this argument, they would need to weigh the risk of a further rise in yields, including the risk of fragmentation, which might then call for buying bonds under the Transmission Protection Instrument (TPI). 

TPI, however, is intended to address a different problem: of yields on some sovereign bonds deviating excessively from that country’s economic fundamentals. TPI was created in response to the 2022 pandemic shock, to enable the ECB to begin tightening while safeguarding against the risk of fragmentation. It would not be the first recourse for a generalised increase in yields. And, as it has not yet been tested, it is unclear how the ECB will determine levels of yields not warranted by fundamentals. It might also not be able to enforce fiscal conditionality, particularly the requirement that to be eligible, a country would have to abide by the EU’s fiscal rules.

Policy messages

The ECB should acknowledge that financial conditions have tightened as a consequence of the Iran-related energy shock. While it is still uncertain how long supply disruptions will last, the bank could emphasise that longer-term inflation expectations remain well anchored and policy rates will continue to address price-stability objectives.

The ECB should indicate that it will consider reducing the pace of QT to prevent excessive tightening of financial conditions. It should remind markets that the pace of QT is flexible and it is not committed to a predetermined schedule. Such a statement would help contain the rise in yields.

A credible communication strategy would comprise the following: the slowdown in QT is partial, the ECB will still unwind but at a slower pace, the slowdown will be temporary to alleviate unnecessary tightening (part of the tightening clearly reflects higher expected inflation and expectations of higher policy rates) and the ECB does not target yield levels.

It could go further by communicating that the forthcoming review of its operational framework for monetary policy, which the ECB has said will happen in 2026, will take into account the energy shock, particularly the impact on market liquidity. This would provide assurance that the supply of reserves will not be constrained.

Source : Bruegel