Isabel Schnabel: Looking back at PEPP implementation since the end of reinvestments

29 April 2025

By Imène Rahmouni-Rousseau and Isabel Schnabel^[1]

In March 2020 the ECB launched the PEPP to counter the serious risks to the monetary policy transmission mechanism and the outlook for the euro area posed by the COVID-19 pandemic.

At the end of 2024 reinvestments under the PEPP came to an end. Over a period of nearly five years the Eurosystem conducted more than 110,000 transactions in private and public sector securities, with portfolio holdings reaching a peak of €1.7 trillion in March 2022.

In January 2025 the Eurosystem further increased the programme’s transparency by offering more backward and forward-looking data and by aligning the PEPP data it publishes with its publication of data for the asset purchase programme (APP).

This blog post looks back through the lens of the enriched data, taking stock and highlighting elements of the programme’s implementation. It is a follow-up to the previous blog published in February 2024 titled “The dynamics of PEPP reinvestments”.^[2]

Balancing transparency and effectiveness

In the past, implementation of the PEPP required us to strike a balance between providing appropriate transparency and safeguarding the effective use of flexibility of purchases over time, across asset classes and among jurisdictions. The granularity, timing and frequency of data disclosure were important dimensions of this balance.

Data disclosure can carry risks for the effective implementation of an asset purchase programme.

High-frequency disclosures, or disclosures without a sufficient lag between operations and publication, could reveal details about the central bank’s implementation strategy, possibly enabling market participants to undermine the effectiveness of the programme.

At the same time, the programme must be transparent so as to inform market participants about the general principles of implementation, ensure accountability and enable an ex-post public evaluation of the effectiveness of the central bank’s actions.^[3] The clear communication on the PEPP’s objectives and features helped to reduce uncertainty and support market functioning.

To strike an appropriate balance, the Governing Council decided in April 2020 to broadly align the granularity of PEPP data with that of the APP. However, it set a bi-monthly publication frequency for granular data of Eurosystem holdings per asset class and per country rather than the monthly frequency set for the APP.

The recent end of reinvestments under the PEPP changes the balance of the above trade-offs, with public accountability now outweighing previous implementation concerns.

Consequently, on 5 December 2024 the Governing Council approved changes to the publication of PEPP data to fully align the level of transparency of the PEPP with that of the APP. Data that was previously available at bi-monthly frequency has been published retroactively at monthly frequency^[4] and, for the first time, we have published both backward and forward-looking redemption data for the PEPP, in line with those for the APP.

The implementation of the PEPP in the light of the new data

The new data allow us to look in more detail at the recent partial reinvestment phase (July to December 2024). Over that period the Governing Council reduced the PEPP portfolio by €7.5 billion per month on average. Chart 1 shows how reinvestments were conducted in 2024, including during the partial reinvestment phase.

Importantly, as noted in our previous blog post, the amount of portfolio redemptions can vary greatly from month to month. Therefore, the Governing Council decided to set the portfolio reduction pace in average terms over July to December, rather than setting an exact monthly pace.

This averaging provided some leeway in how the reduction was implemented. The aim was to balance the Eurosystem’s market presence, taking seasonal patterns of redemptions and issuance into account. In this context, the portfolio reduction pace (grey bars) was adjusted in two months (down in September and up in October), with the average reduction pace amounting to €7.5 billion over the partial reinvestment period.

The data also enable us to revisit episodes of the PEPP when flexibility was actively used. Flexibility was an integral feature of the PEPP, in both the net purchase and reinvestment phases, so we were able to conduct asset purchases flexibly over time, across asset classes and among jurisdictions.

For example, flexibility across asset classes was used to adapt the share of supranational bond purchases to the liquidity available in that market segment.^[5] Flexibility across jurisdictions meant that purchases could deviate temporarily from the capital key of the Eurosystem national central banks. However, the capital key remained the benchmark for the allocation of public sector purchases under the PEPP.

Chart 2 uses the monthly data to illustrate how flexibility across jurisdictions was used to counter risks to the monetary policy transmission mechanism. The vast majority of purchases during the net asset purchase phase were conducted in line with the capital key (dark blue bars).

The chart highlights two episodes where purchases were intentionally allowed to deviate from the capital key. The first was at the start of the programme in 2020, when public sector purchases were increased in jurisdictions most affected by pandemic-related fragmentation risks (first red circle). This led to a modest increase in the cumulative deviation from the capital key (green line), which subsequently receded quickly as financial market conditions stabilised.

The second episode was during the reinvestment phase in the summer of 2022, when fragmentation risks re-emerged. The Governing Council decided to apply flexibility in reinvestments (second circle) with a view to preserving the functioning of the monetary policy transmission mechanism around the time of its first interest rate hike of the current monetary policy cycle.^[6] The fragmentation risks subsided after the Transmission Protection Instrument (TPI) was established, so there was no longer a need to actively use flexibility. The Eurosystem nonetheless maintained its readiness to do so until the end of the reinvestment phase.

As the yellow bars in the chart show, there were small deviations from the capital key throughout the implementation of the programme. As explained in our previous blog post, these reflected technical factors, such as double-smoothing and a scarcity of eligible securities in some jurisdictions.

Looking ahead

The PEPP portfolio is declining at a measured and predictable pace as the Eurosystem no longer reinvests principal payments from maturing securities. The Eurosystem has now started to publish expected redemptions over the following 24 months, at an annual frequency.

This new information, together with the information already being released for the APP, will help market participants to forecast the rundown of Eurosystem monetary policy portfolios more precisely.

Before the data were published, analyst expectations for the redemptions were available in the regular ECB Survey of Monetary Analysts. Chart 3 shows that responses in the January survey fully incorporate the newly published information, thus reducing the uncertainty about this major driver of the reduction of the Eurosystem balance sheet.^[7]

Conclusion

The implementation of the PEPP proceeded smoothly over the entire period of active purchases. While flexibility was an essential feature of the programme, it was used only rarely, and deviations from the Eurosystem capital key swiftly receded. Increasing the transparency of the data on redemptions helps to ensure our accountability, while allowing market participants to forecast the rundown of the PEPP portfolio as they can for the APP portfolio.

Check out The ECB Blog and subscribe for future posts.