Structuring a DORA ICT Third-Party Risk Register for FCA and ECB Scrutiny
A practical guide to building a DORA Article 28 register of information that holds up under parallel FCA and ECB review while giving supervisors confidence your concentration risk is understood and managed. After reading, you will know how to sequence data collection, resolve classification conflicts, and present concentration exposures credibly.
The register of information is no longer a data exercise. For firms in scope of DORA with UK operations and euro area entities, it is the single artefact through which two supervisors will form a view on whether you actually understand your ICT third-party dependencies. Get the structure wrong and you create two problems at once: inconsistent submissions across jurisdictions, and concentration findings that were avoidable with better classification discipline upstream.
Start with the entity perimeter, not the contracts
Most register builds begin with a contract extract from procurement. That is the wrong starting point. Begin with the legal entity map: which group entities are in scope of DORA directly (euro area regulated entities), which are in scope indirectly through intragroup arrangements, and which UK entities fall under FCA operational resilience and SYSC 8 but not DORA itself. The FCA has been clear that it expects UK firms to treat DORA alignment as good practice where it supports PS21/3 outcomes, but the regulatory hook is different. Your register needs to make that distinction visible.
Where a single ICT service supports both a euro area entity and a UK entity, record it once at the service level and tag the consuming entities. Duplicating entries to satisfy both regimes produces inconsistencies that supervisors notice.
Get the function criticality judgement right
The single biggest driver of concentration findings is weak logic on what supports a critical or important function (CIF). Firms either over-scope, flooding the register with low-materiality services and diluting the signal, or under-scope, which looks like minimisation when the ECB benchmarks across peers.
Good practice: tie CIF designation explicitly to the impact tolerances set under your operational resilience framework. If a service supports a function with a defined impact tolerance, it is in. If not, document why. The ECB's SREP teams and the FCA's operational resilience supervisors will both look for this linkage. When it is absent, they assume the classification is procurement-driven rather than risk-driven.
Build the concentration analysis before the register is final
Concentration risk findings are rarely about the raw numbers. They are about whether the firm has done the analysis and can show it. Before submission, run three cuts of the data:
- Provider concentration: total CIF-supporting spend and service count per provider, flagging any provider supporting more than one CIF.
- Geographic and sub-contractor concentration: particularly fourth-party dependencies on hyperscalers, which both supervisors are actively mapping across the sector.
- Substitutability: for each CIF-supporting arrangement, a documented assessment of exit time, portability, and tested alternatives.
Where concentration exists, say so plainly in the accompanying narrative and set out the mitigation: contractual exit rights, tested portability, stressed exit scenarios, board-level risk acceptance where relevant. Supervisors respond badly to registers that present concentrated exposures as diversified through creative categorisation. They respond well to honest concentration paired with credible management.
Reconcile the FCA and ECB lenses explicitly
The ECB will focus on the Article 28 data fields, subcontracting chains, and whether the register supports the Oversight Framework for critical third-party providers. The FCA will focus on whether the register evidences the firm's own understanding of important business services and impact tolerances, consistent with PS21/3 and the forthcoming CTP regime under FSMA 2023.
Build one register, then produce two short cover notes. The ECB note should address data completeness, subcontractor visibility, and group-level concentration. The FCA note should address linkage to important business services, severe but plausible scenario testing, and board oversight. Same underlying truth, two framings.
What good looks like at submission
A register where every CIF-supporting arrangement has a named accountable SMF or equivalent, a documented substitutability assessment, a tested exit plan where material, and a board or committee minute reference showing the concentration position has been actively considered. If you cannot produce those four things for your top ten arrangements by provider spend, you are not ready to submit.
Next step
Before the next submission cycle, commission a dry run against your current register using the three concentration cuts above. The gaps you find now are findings you prevent later.
Frequently Asked Questions
Should we submit the same register to the FCA and ECB?
The formal Article 28 register goes to your lead DORA competent authority routing to the ESAs. The FCA does not receive the DORA register directly, but will ask for equivalent information through operational resilience supervision and, in time, the CTP regime. Build one dataset that can serve both.
How do we handle intragroup ICT arrangements?
They are in scope and must be recorded. Supervisors are particularly alert to intragroup arrangements being used to obscure external concentration. Record the ultimate external provider where the intragroup entity is a conduit.
What triggers a concentration risk finding in practice?
Usually one of three things: a single provider supporting multiple CIFs without documented substitutability, hyperscaler dependencies not identified at the fourth-party level, or exit plans that have never been tested. The finding is almost always about the absence of analysis, not the exposure itself.
How detailed should the subcontractor chain be?
Down to any subcontractor that materially supports a CIF. 'Materially' is a judgement call, but the test is whether failure of that subcontractor would breach an impact tolerance. If yes, it goes in the register with the same discipline as a direct provider.
Who should own the register internally?
Operational resilience or ICT risk, with procurement as a data contributor and the accountable SMF (typically SMF24 in the UK) signing off. Procurement-owned registers consistently produce weaker criticality judgements.
Frequently asked questions
Should we submit the same register to the FCA and ECB?
The formal Article 28 register goes to your lead DORA competent authority routing to the ESAs. The FCA does not receive the DORA register directly, but will ask for equivalent information through operational resilience supervision and, in time, the CTP regime. Build one dataset that can serve both.
How do we handle intragroup ICT arrangements?
They are in scope and must be recorded. Supervisors are particularly alert to intragroup arrangements being used to obscure external concentration. Record the ultimate external provider where the intragroup entity is a conduit.
What triggers a concentration risk finding in practice?
Usually one of three things: a single provider supporting multiple CIFs without documented substitutability, hyperscaler dependencies not identified at the fourth-party level, or exit plans that have never been tested. The finding is almost always about the absence of analysis, not the exposure itself.
How detailed should the subcontractor chain be?
Down to any subcontractor that materially supports a CIF. 'Materially' is a judgement call, but the test is whether failure of that subcontractor would breach an impact tolerance. If yes, it goes in the register with the same discipline as a direct provider.
Who should own the register internally?
Operational resilience or ICT risk, with procurement as a data contributor and the accountable SMF (typically SMF24 in the UK) signing off. Procurement-owned registers consistently produce weaker criticality judgements.
Related guides
How to Identify What Will Give a Regulator Confidence in a Proposed Change
This guide explains how to work out what a regulator actually needs to see before they will support a proposed change, from evidence to governance to post-implementation controls. After reading, you will be able to build a submission and engagement plan that reflects how supervisors actually form judgements.
How to Structure a Section 166 Scoping Response That Limits Reviewer Overreach
This guide sets out how to respond to an FCA or PRA Section 166 scoping notice in a way that constrains the skilled person's remit without antagonising the regulator. After reading, you will know how to shape the scope, methodology, and reporting terms before the skilled person is appointed.
How to Structure a Section 178 Change in Control Notification That Holds Up
A practical guide to assembling a Section 178 notification that the PRA and FCA can assess on first pass, without multiple rounds of clarification. Covers sequencing, evidence quality, and the judgement calls that determine whether the 60 working day clock runs clean.
Structuring a Dear CEO Letter Response That Lands Credibly
This guide sets out how to construct a Dear CEO letter response that answers the FCA's concerns with evidence, candour, and proportionate commitments. After reading it, you will know how to sequence the response, calibrate tone, and close the loop in a way that builds supervisory trust rather than prompting further questions.
Structuring a Section 167/168 Investigation Response Strategy
A practical guide to building an investigation response that protects the firm's legal and commercial interests while demonstrating genuine cooperation with the FCA. Readers will finish with a clear framework for governance, document handling, witness preparation, and stakeholder sequencing in the first critical weeks of an investigation.
Where internal confidence may exceed external evidence
Polar Insight helps leadership teams test critical assumptions against stakeholder, market, regulatory, and operational reality before risk compounds.
Explore Stakeholder Proximity