Response to consultation on Regulatory Technical Standards on on the threshold of activity at which Central Securities Depositories (CSDs) providing ‘banking-type ancillary services’
Q1. Do you agree with the proposed approach for determining the threshold referred to in Art. 54(5) of the CSDR?
Our responses to this consultation are informed by a risk accounting-based perspective that prioritizes the real-time quantification of residual non-financial risk at the transaction level. This approach, developed through academic collaboration and practical fieldwork, aims to provide regulators and institutions with objective, consistent, and forward-looking visibility into risk accumulation. A detailed explanation of this methodology is presented in the vision document attached to our submission. It outlines how residual risk quantification using standardized Risk Units (RUs) can enhance current supervisory practices by improving systemic risk detection, tiering accuracy, and proportionality in regulatory requirements.
We support the principle of setting a threshold to identify credit institutions that should be subject to enhanced oversight when acting as settlement agents. However, we would like to raise a concern regarding the reliance on volume and value metrics alone to determine systemic relevance.
The proposed methodology defines a core threshold based on historic settlement activity (measured in volume and value terms), and then introduces two tiers: one for institutions "just above" the threshold and another for those "well above" it. While we acknowledge the rationale for proportional application of prudential requirements, we note that this approach depends entirely on post-factum measurements — that is, activity data gathered over a rolling 12-month period. As a result, systemic risk may accumulate unnoticed before an institution crosses the threshold, and designation can lag behind actual risk exposure.
Furthermore, volume and value are not necessarily correlated with risk intensity. An institution settling a lower number of complex, high-risk transactions (e.g., involving FX, derivatives, or collateral dependencies) may pose greater operational and systemic risk than one processing high volumes of standard equity transactions. Therefore, while activity metrics are a practical starting point, they do not provide a complete picture of emerging risk.
We suggest that future iterations of the threshold methodology consider integrating an objective, risk-based measure — such as the quantification of residual non-financial risk using standardized units (e.g., Risk Units or RUs). Such an approach would enable earlier detection of systemic exposures and better alignment between designation and true risk accumulation.
Alternative regulatory choice: We propose a two-layer approach. The current threshold methodology could remain as a screening tool, but institutions identified as threshold-adjacent or above should also be required to implement a standardized, transaction-level risk measurement framework. This would provide supervisors with a dynamic, forward-looking view of risk accumulation, rather than relying solely on backward-looking transactional data. By quantifying risk as it emerges — at the transaction level and in near real-time — Risk Accounting enables proactive risk mitigation and improves the timeliness of supervisory responses. This could materially reduce systemic exposures before they escalate to threshold breaches.
Q2. Do you think that other elements should be taken into account in the proposed approach? If yes, which ones?
Yes, several additional elements should be considered to better capture the systemic relevance of credit institutions acting as settlement agents under Article 54(5) of the CSDR.
First, we recommend incorporating the risk sensitivity of transaction types. The current approach assumes equivalence between different types of settlement activity based on volume and value, but transactions involving derivatives, cross-border elements, or FX-linked components carry materially higher operational and systemic risks. Risk weighting by transaction type would create a more accurate profile of actual risk exposure.
Second, the nature of the institution’s client base should be considered. A credit institution servicing systemically important financial market infrastructures (FMIs) or large institutional clients introduces concentrated risk even at lower activity levels. This could be a factor in assessing "substantial importance" as intended in Article 54(5).
Third, cross-border operational complexity and settlement in multiple currencies should be included as a risk factor. These elements introduce elevated intraday liquidity, operational, and legal risks that are not visible through aggregate activity metrics.
Fourth, institutions should be required to report stress-indicative metrics such as settlement fails, margin or collateral shortfalls, and late-day liquidity compression. These forward-looking indicators could act as early warning signals even before volume-based thresholds are crossed.
Finally, we propose introducing a standardized residual risk measure (e.g., Risk Units) across settlement activities. This would enable proportionality and consistency in designation by providing a risk-adjusted complement to volume-based thresholds.
Alternative regulatory choice: The EBA could implement a multi-factor designation framework that integrates transaction complexity, client criticality, geographic scope, and operational resilience, alongside transaction volumes and values. This would strengthen alignment with Article 54(5)’s emphasis on "substantial importance" to the functioning of securities settlement systems.
Q3. Do you agree with the proposed levels set out in the proposed approach for the different parameters?
We recognize the effort to calibrate threshold levels using available historical data to identify credit institutions of systemic importance. However, we are concerned that the levels set out in the proposed approach may not fully reflect the underlying risk exposure that Article 54(5) aims to address.
Because the levels are based solely on quantitative metrics such as settlement volume and value, they may result in either overinclusion of low-risk institutions or underinclusion of high-risk institutions whose transaction types or operational structures present significant systemic risk despite lower volumes.
Moreover, uniform thresholds do not account for market diversity. For instance, institutions active in cross-border or multi-currency settlements may face operational and liquidity risks disproportionate to their transaction volumes. Similarly, firms with concentrated exposure to FMIs or specific client types may fall below the numeric threshold while presenting substantial systemic impact.
We recommend that threshold levels be supplemented with a calibration mechanism that includes risk-intensity adjustments. This could be achieved by integrating a weighting system based on transaction type, operational complexity, or the presence of stress indicators. A complementary risk-adjusted exposure measure (e.g., residual Risk Units) could help refine threshold applications and provide more dynamic responsiveness to market structure changes.
Alternative regulatory choice: In addition to fixed numeric levels, the EBA could introduce a hybrid calibration model, whereby threshold triggers are stress-tested against qualitative and risk-weighted variables, with supervisory discretion to designate institutions whose structural features suggest material systemic relevance even if they fall marginally below the quantitative thresholds.
Q4. Do you agree with the proposed basic risk management and prudential requirements? If no, please provide rationale and an alternative proposal.
We agree with the inclusion of foundational governance, risk identification, and operational resilience measures as part of the basic requirements. These are in line with good risk management practices and provide a necessary baseline for institutions just above the CSDR threshold.
However, we would express a concern regarding the assumption that structural and procedural compliance alone is sufficient to manage systemic risk. The current approach focuses on the presence of controls — internal audit, business continuity, and liquidity monitoring — but does not require institutions to measure or report the actual residual risk they carry through their settlement operations.
This creates a situation analogous to designing a dam based on theoretical safety specifications, without knowing — or measuring — the volume or speed of water building up behind it. Without an objective and consistent mechanism to quantify risk accumulation (e.g., using residual Risk Units), supervisors and institutions alike lack the real-time visibility needed to manage and preempt emerging systemic exposures.
We recommend enhancing the basic requirements by introducing a standardized, activity-linked risk quantification method that reflects the actual volume, complexity, and control effectiveness of settlement-related operations. This would enable institutions to not only design robust controls but also to validate their sufficiency against measurable risk levels.
Alternative regulatory choice: The EBA could amend the basic requirements to include a requirement for transaction-level residual risk quantification, expressed using a standardized metric such as RUs. This would improve comparability across institutions, reinforce proportionality, and provide supervisors with actionable insights beyond structural checklists.
Q5. Do you agree with the proposed level of settlement activity, which determines whether only basic or both basic and advanced risk management and prudential requirements are applied?
We support the principle of applying risk management and prudential requirements proportionately to the level of settlement activity. Differentiating between institutions just above the threshold and those significantly above it aligns with Article 54(5)'s intent to focus regulatory intensity where systemic importance is greatest.
However, the current approach relies exclusively on retrospective settlement activity data (volume and value) to determine tiering. This may lead to misclassification, especially for institutions experiencing rapid growth, accumulating high-risk exposures in less visible ways, or servicing particularly sensitive counterparties or infrastructures.
We believe that volume-based activity measures should be complemented by real-time residual risk quantification, which captures the true intensity of risk exposure across different transaction types and operational contexts. This would allow the regulatory framework to more accurately reflect systemic relevance and escalate supervisory requirements before latent risks materialize.
Alternative regulatory choice: We propose introducing a dynamic tiering mechanism that combines transaction activity with real-time risk exposure metrics. Institutions could be transitioned from basic to advanced requirement status not solely based on trailing 12-month settlement activity, but also based on accumulating residual risk indicators. This would create a more responsive and preventive prudential regime.
Q6. Do you agree with the proposed advanced risk management and prudential requirements? If no, please provide rationale and an alternative proposal.
We support the inclusion of advanced requirements for institutions whose scale and systemic relevance warrant enhanced supervisory scrutiny. Measures such as dedicated risk governance structures, stress testing, intraday liquidity risk monitoring, and recovery planning are necessary to mitigate the broader financial stability risks posed by large-volume settlement operations.
However, we believe the proposed framework still lacks a quantitative dimension that reflects the actual level of risk being accepted and accumulated. While structural requirements define oversight and control procedures, they do not address the need for measurable, real-time indicators of risk exposure build-up.
Advanced institutions — by definition — operate at a level where qualitative governance alone is insufficient. The absence of a standardized, real-time residual risk quantification approach leaves a significant visibility gap for both the institution and the supervisor. Without objective, consistent, transaction-level metrics (e.g., residual Risk Units), it is difficult to evaluate the effectiveness of the advanced controls or to intervene proactively when systemic exposures begin to form.
We therefore recommend that the advanced prudential regime require institutions to implement a risk accounting system that quantifies and tracks residual risk across their settlement activities. This would not only enhance stress testing and liquidity risk assessments, but also provide supervisors with timely, comparable, and actionable insights into the evolving risk landscape.
Alternative regulatory choice: Extend the advanced requirements to include mandatory real-time residual risk quantification, with periodic reporting to supervisors. This would provide an empirical foundation for supervisory decision-making, improve early warning capabilities, and ensure that institutions designated as systemically important demonstrate both qualitative governance and quantitative transparency over their settlement-related risk exposures.