Strengthening the supervisory grip on securitisation: reading the data, anticipating the risks

Keynote speech by Pedro Machado, Member of the Supervisory Board of the ECB, at the PCS Symposium “Securitisation – adapting to the new era”

Lisbon, 14 May 2026

I am delighted to be here with you today.

This will be my third in a series of speeches on securitisation.

I gave the first in September 2025 and the second in March 2026.^[1] The former provided a comprehensive overview of market developments for securitisation in the EU as seen from a supervisory perspective, while also examining what could be done to adjust the regulatory framework.^[2] The latter focused on synthetic securitisations, drawing some lessons from recent research on the topic and providing the ECB’s perspective on the growth of this particular market segment in recent years.^[3]

In today’s speech, by drawing on the latest data available to ECB Banking Supervision, I would like to substantiate the assessment of the risks we see in significant risk transfer securitisations and, in particular, synthetic securitisations.

A constructive reading of EU market developments

For an evidence-based analysis, it makes sense to start with some solid numbers on market trends.

In my previous speeches I referred to the rapid growth of synthetic securitisation issuances. The last year has confirmed the overall trend observed in previous years. That is, total issuances of synthetic securitisations by significant institutions reached a total of €258 billion in 2025, as measured by the notional amounts of the securitised portfolios.

This represents an increase of 47% year on year for synthetics, with a total securitised amount of €258 billion in 2025 compared with €175 billion in 2024. And if we look at the growth between 2022 and 2025, the increase amounts to 90% in just a few years – this shows the speed at which the synthetic market has been developing.^[4] In terms of stock, we observe a similar trend: the stock of underlying exposures in securitisations originated by significant institutions has risen from €223 billion at the end of 2022 to €570 billion at the end of 2025.

Let me clarify something about traditional securitisations. The numbers above do not include traditional securitisations, including non-performing exposures, which – until 2021 – were driving issuance volumes up before starting to decline.^[5] Today performing traditional securitisations represent only a fraction of synthetic issuances, and issuances remain concentrated – coming from only a handful of banks. This contrasts with the synthetic market where we have seen three significant groups issuing their first transactions in 2025.

Overall, these developments support the EU’s goal of reviving securitisation markets – an objective that led the European Commission to propose amendments to the regulatory framework for securitisation. I would like to emphasise that the ECB has been supportive of the proposals and recognises that a well-structured securitisation framework can, in principle, support economic growth. However, this positive outcome is not a given and will always depend on whether banks channel loans to the real economy following the capital relief they derive from their risk transfer securitisations.^[6]

If we look back at previous regulatory initiatives on securitisation, judging from the increase in issuance volumes until now, the 2021 introduction of the new framework for simple, transparent and standardised securitisations seems to have been very effective.

So you could assume that adopting the legislative package – which could still happen this year – may further fuel the growth of synthetic issuances. the EU co-legislators’ eventual adoption of the proposal to recalibrate capital requirements for senior securitisation positions held by credit institutions might lead to risk weights that would, in some cases, be below those applicable during the great financial crisis, as mentioned in the relevant ECB opinion.^[7] However, there are several elements, some of them more structural, some of them linked to the current economic situation, that might slow down or limit the growth potential.

Among the structural factors, I would like to mention a regulatory backstop, such as the leverage ratio, which does not recognise any capital relief from synthetic securitisation, and some banks could be constrained by this rather than the risk-weighted asset framework.^[8] Furthermore, banks’ risk appetite and capital planning can be a factor where banks need to manage the possible refinancing risk well if the synthetic transaction either becomes more expensive or cannot be refinanced, meaning the capital relief is cancelled in the future. Additionally, the creditworthiness assessment of their investors and rating agencies, might contribute to limiting the absolute volume of synthetics that can be issued. Furthermore, when volume is growing substantially, market absorbency capacity might function as an additional factor limiting the growth potential of synthetics.

Market conditions, notably the pricing of risks by investors, also play an important role in yearly issuance volumes. We observed a notable – albeit temporary – decline in synthetic issuances of around 6% in 2023, which corresponds to an increase in the average cost of protection across all banks for the same year. However, as seen over a longer period, the average cost of protection for significant institutions has steadily decreased between 2020 and 2025. It is possible to infer that this shift has facilitated the market growth we have observed over the same period.

I will now say a few words about our take on market developments over the first months of 2026, for which the ECB has not yet published data. It would seem that issuances have been on a downward trend compared with the same period in 2025. I would expect the complex geopolitical environment to have played a strong role here, with both originators and investors adopting a wait-and-see approach and possibly postponing issuance. Increasing cost of protection, in particular, may also have convinced originators not to rush to the market and instead to take a bit more time.

In January 2026 the ECB introduced fast-track procedures for banks’ applications to reduce their own funds and for significant risk transfer securitisations. This enables the ECB to respond more quickly to the more standardised transactions, while maintaining supervisory rigour and freeing up internal resources to look at the more complex transactions. Needless to say, this new fast-track procedure does not change applicable prudential requirements.

During the first four months of 2026 significant risk transfer securitisation volumes remained modest: only two transactions were submitted to the ECB via the fast-track process. The ECB completed its assessment of both transactions in just eight working days, whereas the previous process typically took several months. Looking ahead, the number of notifications is expected to rise in line with the higher volumes planned to be issued and as originators become more familiar with the new fast-track procedure for significant risk transfer securitisations.^[9]

While it is too early to draw clear conclusions, we note a positive dynamic around the new fast-track process. Of course, we will also have to see how our fast-track process fits into the new regulatory environment once the legislative package has been finalised. I would now like to share with you some of our main conclusions from our overview of risks.

What the data can tell us as supervisors

In recent months, in keeping with our evidence-based approach, we have increased our supervisory activities to complete our overview of the risks from securitisations that the banks we supervise might be exposed to. This has required substantial efforts from supervisors to improve the quality of data we collect. It is also certainly an area where further work will be needed, both from supervisors and from banks. We must collectively ensure that the information being reported is consistent, coherent and comparable over time and among banks.

On the overview of risks, we have looked at those arising from synthetic securitisations, namely: (i) rollover risk, i.e. the risk that an originator would not be able to renew its risk transfer securitisations once the capital benefit diminishes over time, either due to the natural amortisation of the securitised portfolio or because the originator decides to call the transaction early; (ii) counterparty credit risk, i.e. the risk that the protection provider could fail to meet its contractual obligations before the final settlement of the securitisation in unfunded synthetic structures; and (iii) flowback risk, i.e. the risk that the capital requirements from the retained senior tranches could increase sharply at some point during the life of the securitisation.

When looking at these risks, we need to consider the impact of certain risk amplifiers, such as any investor concentration at market level and bank level. We also need to consider potential sources of interlinkages between banks and non-bank financial institutions, which could undermine the effectiveness of the risk transfer – for example, the financing provided by banks to investors in risk transfer securitisations and collateralised by credit linked notes.

So, what can we conclude? Overall, current securitisation risks appear to be manageable overall, at both bank level and system level – but of course further growth could change this assessment. Let me zoom in concisely for each of the risk categories we are monitoring.

For rollover risk in general, we did not detect any “maturity wall”, i.e. no material concentration of maturities, and capital relief rollover does not generally seem to pose a larger risk to banks’ capital positions. This does not mean that we did not detect individual situations where further attention might be required.

For counterparty credit risk, we have mainly focused on unfunded protection from counterparties other than governments and development banks. Overall, unfunded protection provided by those investors only represents 11% of all outstanding protected tranches, while the protection is collateralised for most of the tranches. We have noted, however, that this market segment is rather concentrated, with just a few subsets of investors having a substantial share. We also note that the capital calculated for these repo positions seem to differ from bank to bank, and this is something that we will continue to examine.

Lastly, we note that synthetics are generally more prone to flowback risk. This is because the senior tranche is often fully retained by the originator. This differs from, for instance, traditional securitisations. The primary mitigant against flowback risk is the robustness of the risk transfer, which involves meeting supervisory expectations concerning the level of credit enhancement for protecting the senior tranche. For each transaction, this is evaluated on a case-by-case basis during the significant risk transfer assessment conducted by the supervisor. The possible impact of flowback risk at bank level also needs to be considered. This requires an adequate stress-testing framework to quantify the risk of additional losses.

While this overview of risks has not revealed an immediate, larger cause for concern, we intend to continue our work, notably on investor concentration and stress-testing capabilities, and improve data quality. It is particularly important that we can keep up the pace of market growth, which could further accelerate following the adoption of regulatory changes later this year.

In particular, there are two factors that should be considered carefully.

The first is the pace of growth itself. Volumes are rising quickly, and when this happens the interconnections between banks and the non-bank financial sector deepen in ways that are not always fully mapped. The counterparties to synthetic risk transfer transactions are, by and large, non-bank financial institutions, such as private credit funds, hedge funds and – for unfunded credit protection – insurers. This creates channels through which risk can flow back to the banking system, whether through funding linkages, operational dependencies or the return of exposures under certain stress scenarios. As these linkages deepen, even risks that might once appear moderate can take on greater significance.

The second factor is the quality of the data. Quite plainly, data gaps and issues with data quality continue to impede a fully reliable assessment of spillover risks. This is not criticism of any specific institution but rather a structural observation about the current state of the reporting landscape. We have made progress, but further efforts are needed, especially with regard to the assessment of exposures, the characteristics of investor bases and the post-transaction performance of reference portfolios. Without this information, supervisors and market participants alike are navigating with only partial visibility. That is not a sustainable position for a market of this scale.

All of this leads to a simple conclusion. We need to be proactive as risks develop, rather than simply reacting once they have crystallised. This necessitates improving supervisory oversight now, while the market is functioning well; it means enhancing how we monitor individual transactions; and it calls for closing existing gaps in the data. It is far easier to set standards when the sun is shining – the task becomes much more challenging once the storm arrives.

A call for simplicity and standardisation

As I have mentioned in my previous speeches, we prefer simplicity and standardisation. We made it clear in the ECB opinion from November 2025 that the prudential and regulatory framework should encourage the establishment of simple, standardised transactions – it should discourage complex transactions that provide limited benefits for the financing of the economy and that pose higher risks for credit institutions acting as investors and originators.^[10] We want to avoid a situation where the growth of the market is accompanied by greater use of complex structural features and where banks seek to increasingly securitise exotic and opaque asset classes without proper risk management.

First, sound market development is conditional on the ability to attract new market players, in particular new investors – to match growing issuance volumes from banks. Simplicity is key, as it will allow new entrants to perform due diligence more easily. This is particularly important as the legislative changes currently under discussion will also make the due diligence requirements for investors more proportionate. Simplicity and standardisation are also beneficial to new originators, as they usually have less advanced risk management frameworks.

Conversely, complexity would likely deter new investors. It also requires banks to update their risk management frameworks to match the complexity of their transactions.

Complexity also puts a drag on supervisors, especially when we are confronted with innovative transactions for which the regulatory framework is not clear, or simply silent. This requires a substantial amount of work for single transactions, which is not always matched by clear benefits for the financing of the economy. An example of this would be the securitisation of counterparty credit risks, which presents specific characteristics that the securitisation framework cannot capture accurately, as it is best suited to plain credit risk.

Governance must match growth

It is clear to us as supervisors that risk transfer securitisation is increasingly becoming a capital-management tool used by increasing numbers of banks. More banks are incorporating significant risk transfer issuance explicitly into their capital trajectories. It is argued that properly structured significant risk transfers are a legitimate and efficient instrument for managing credit risk. But when it is used as a capital relief measure supporting excessive payouts and this outpaces the development of the relevant risk management frameworks, banks can be exposed to risks that they cannot fully understand or measure. This mismatch between financial engineering and governance is a vitally important issue that warrants close observation going forward.

Long-standing issuers have established securitisation programmes that they roll out every year. We need to ensure that their risk management capabilities stay up to date as issuance volumes gradually increase. And as for the newcomers to the world of significant risk transfers, we need to ensure that those banks fully understand the risks arising from securitisation.

With more securitisation comes a heightened need for robust governance and risk management. Banks’ risk management frameworks for risk transfer securitisation should therefore be fit for purpose and capable of adequately supporting banks’ overall securitisation strategies.

Conclusion

Let me offer some concluding remarks.

The data shared today paints a clear picture. The European significant risk transfer market, particularly the synthetic segment, has grown rapidly. We can observe a 47% increase in yearly issuance volumes between 2024 and 2025, and a near doubling since 2022. The pool of issuing banks is widening, investor interest remains robust and the legislative package currently being discussed should support further market development overall.

The supervisory side is more nuanced. For now, risks at both bank level and system level currently seem to be manageable. We have not detected a “maturity wall”, an acute concentration of unfunded counterparty exposures or an immediate flowback vulnerability. But we do see volumes rising quickly and deepening interlinkages with the non-bank financial sector, as well as persistent data gaps and instances of risk management frameworks that have not yet caught up with the scale and sophistication of their issuance activity.

Against this backdrop, our responsibility as supervisors is manifold.

First, we need to continue investing in data and analytics so that supervisors and market participants can have a clear view of this market. The reconciliation exercise we have completed in recent months has improved the data I have referred to today, but there is still much to do.

Second, we need to ensure that banks’ governance and risk management practices evolve in parallel with the growth of their issuance activity.

Third, we need to preserve and – where appropriate – refine the supervisory architecture that has served European significant risk transfers well. The ex ante, case-by-case assessment of significant risk transfer applications is a distinctive feature of our framework, which contributes meaningfully to the quality and credibility of European securitisation transactions. The fast-track process introduced by European banking supervision at the start of this year aims to streamline the approval process, hence it is not a departure from it.

Fourth, we must ensure that market participants remain anchored in the principles of simplicity and standardisation, as these are consistent with a sound supervisory preference for transparency over complexity.

None of this is a one-sided endeavour. A well-functioning market requires good faith, reliable exchange between supervisors and the institutions we supervise; between supervisors and the investors who ultimately bear the risk; and between the supervisors and the policymakers that shape the framework within which we all operate.

The questions raised by the growth of significant risk transfer in Europe – about market structure, about the boundaries between banks and the non-bank financial sector, about the right balance between innovation and prudence – require sustained attention from supervisory authorities and an ongoing, productive dialogue with all stakeholders.