Prepare a filing

Prepare and approve an IRRD recovery plan

Configure solo or group scope, calculate governed scenarios and actions, approve immutable evidence, and respond to operational IRRD triggers.

Version 2026.8.115 minAdministrator · Preparer · Reviewer · Read onlyReviewed 2026-09-02

Who can perform the task

Preparers and administrators create recovery plans, bind approved source facts, set the undertaking's approved warning and breach thresholds, calculate results, and submit a current plan for review. Reviewers and administrators can approve the exact calculation submitted for review only when they did not create or submit that plan. They can also return it for correction with a required comment. Read-only users can inspect plans and calculation evidence available to their organization.

IRRD is a controlled-pilot feature. A platform administrator must enable the named organization before its navigation, pages, actions, background jobs, and notifications become available. Disabling access preserves approved plans and evidence but stops further IRRD access and execution.

IRRD controlled-pilot administration listing named organizations and their current enablement state.

Before you begin

Confirm the reporting entity, plan period, source-data observation date, and the approved source snapshot. Agree each threshold through the undertaking's governance process. Prepare the undertaking profile, governance and escalation arrangements, communication strategy, and an approved supporting document such as the ORSA or liquidity plan. The supporting reference requires its document identifier, revision, and SHA-256 checksum; a title alone is not provenance. The initial catalog covers the indicator families described in EIOPA's IRRD indicator guidance, but its formulas are labelled SolvencyBridge methodology unless a definition explicitly records a regulatory methodology. SolvencyBridge does not invent universal authority thresholds.

Quick path

  1. Confirm that the controlled IRRD pilot is enabled and that you have the required role.
  2. Define the plan scope, recovery indicators, sources, thresholds, and governance owners.
  3. Run the governed calculations, scenarios, and operational feasibility checks.
  4. Review source readiness, assumptions, exceptions, and approval evidence.
  5. Approve the immutable snapshot and monitor subsequent trigger events without treating the product as regulatory advice.

Set scope and proportionality

Record whether the plan covers one undertaking or a group. A solo scope pins the undertaking and reporting currency. A group scope also pins the lead undertaking, every included legal entity, membership dates, approved source snapshot and checksum, reporting currency, FX evidence, and explained elimination adjustments. Additive facts are converted with the pinned rates and summed before ratios are calculated. Ratios are never averaged. A non-additive fact needs an approved group-level source fact.

Group approval fails closed when a member or member fact is missing or stale, an FX rate is absent, periods or units conflict, an elimination lacks evidence, or a required group-level fact is unavailable.

Record proportionality as Undetermined, Standard, or Simplified. For Standard or Simplified, record the criteria, rationale, author, time, and authority evidence when available. SolvencyBridge does not infer simplified- obligation eligibility. Undetermined proportionality cannot cross review.

IRRD scope and proportionality workspace showing the controlled human decision and reporting currency.

Create and calculate a plan

  1. Open IRRD from the authenticated filing portfolio.
  2. Select the reporting entity, name the plan, and enter its period end.
  3. Work through the plan outline in the explorer. Only the selected section is shown, while unsaved field values remain in the browser when you move between sections: Scope and proportionality, Recovery-plan content, Source facts, and Review readiness. The explorer reports completed sections, the next incomplete section, and the last saved time. A section is marked complete only after its required evidence has been saved. The latest calculation, the scenario comparison, and the scenario matrix are tabs in the bottom panel, and the plan's source snapshot, indicator count, and scenario count stay visible in the inspector while you edit.
  4. Complete Recovery-plan content. Describe the undertaking and relevant group context, critical functions and dependencies, plan ownership and escalation, and internal, external, and supervisory communications. Bind one supporting ORSA or liquidity-plan reference to its exact document identifier, revision, and SHA-256 checksum. Saving this content creates a new plan revision; incomplete content can remain in draft but cannot cross review.
IRRD recovery-plan content workspace with undertaking, governance, communication, and pinned document-reference fields.
  1. Select an approved filing snapshot. Map every required fact to exactly one canonical source cell and confirm its unit. SolvencyBridge retains the snapshot, cell, observation date, and checksums in the calculation trace. Manual draft facts can be calculated for preparation, but they cannot be submitted for review and are not written back to a Solvency II filing. The fact list is compact by default; expand one fact at a time to edit its value, unit, canonical cell, observation date, and manual source reference.
  2. Expand Stress scenario, Management action, and Indicators and thresholds only when working on that decision context. All fields remain part of the same explicit save-and-calculate boundary.
  3. Enter the warning and breach threshold approved for each indicator.
  4. Select Save and calculate, then wait for the updated results.
  5. Review every result. Expand Calculation trace to inspect the formula, exact source facts, source checksums, scenario transformations, and confirmed management actions used by the engine.
IRRD source-readiness workspace showing an approved filing snapshot and canonical source-cell mappings.

Compare scenarios and management actions

Each scenario records its family, version, applicability, severity, evaluation horizon, assumptions, provenance, transformations, and package checksum. Combined scenarios remain explicit packages; installing a later package does not alter an approved plan.

For each scenario, calculate the baseline, the scenario before actions, and the scenario after explicitly selected actions. The matrix retains values, deltas, trigger states, completeness, warnings, traces, action order, unavailable actions, and checksums. Repeating the same pinned inputs produces the same matrix.

Before selecting a management action, record its owner, execution window, realization lag, prerequisites, dependencies, mutual exclusions, resource claims, legal and operational limitations, feasibility rationale, and evidence. The preparer must explicitly confirm the deterministic effect. Cycles, unresolved prerequisites, shared-resource conflicts, repeated transformations without sequencing, and effects unavailable by the selected horizon fail closed. SolvencyBridge does not forecast effects or optimize action selection.

Results fail closed as Incomplete, Unsupported, or Unconfigured when required data or executable semantics are unavailable. A Warning or Breach result records the threshold outcome; it is not a recommendation to execute a recovery action.

Review and approval

  1. Complete all recovery-plan content sections and resolve every incomplete, unsupported, or unconfigured calculation. Warning and breach trigger states remain visible for human review; they do not by themselves assert that a management action should be executed.
  2. Add an optional review comment and select Submit for review. This pins the exact plan revision and calculation checksum; subsequent draft editing cannot silently replace it.
  3. An independent reviewer or administrator inspects the inputs, thresholds, results, and traces. The plan creator and submitter cannot approve it.
  4. Select Approve and lock, or add a required comment and select Return for correction. Submission, return, and approval events remain visible in the review history with their actor, time, revision, and checksum.
  5. Download the evidence pack from the approved plan. It contains the plan, regulatory source manifest, scope and proportionality snapshots, calculation request and result, scenario matrix, action feasibility and ordering, approval history, trigger-history capture, source and definition checksums, traces, and explicit limitations. Its own checksum detects later modification. The downloaded pack is the exact evidence approved with the locked plan; it is not rebuilt from later draft changes. Previously approved version-1 packs remain verifiable byte for byte.

Monitor and resolve triggers

Only active approved plan snapshots can generate operational triggers. New approved source data is evaluated against the active approved plan. Draft calculations never create live events.

Open Trigger monitoring to filter warning and breach events. The status filter — open, acknowledged, resolved — sits in the tabs above the queue, and severity, scope, plan, entity, overdue state, and detection time are filters in the explorer. Each event lists its indicator, plan, severity, detection time, and acknowledgement deadline, and breaches and warnings carry a marker on the row. The detail workspace keeps the event status, detection time, deadline, and pinned source in the explorer. It shows the calculation trace and later evaluation link in the work area, with the commented reviewer action beside the evidence on open and acknowledged events. Existing acknowledgement and resolution comments remain visible with the evidence.

IRRD trigger-monitoring queue with governed filters and an explicit empty state.

Warnings are due for acknowledgement within 24 hours and breaches within 4 hours unless an approved organization policy says otherwise. A reviewer or administrator must add a comment to acknowledge an open event, then add a second comment to resolve the acknowledged event. A later normal evaluation does not silently close the event. In-app notification is always recorded; configured webhook delivery retries independently and cannot change trigger state.

Limitations

  • The current product uses the EU baseline and does not claim national IRRD transposition coverage.
  • Controlled transformations do not replace actuarial, liability, claims, asset, capital, or internal-model projections.
  • Approved snapshot binding proves retained content identity and lineage, not source accuracy or the suitability of a value for a recovery-plan purpose.
  • A checksum-pinned document reference proves which external document revision was cited. SolvencyBridge does not infer or validate that document's content.
  • An approved evidence pack records the insurer's SolvencyBridge workflow. It is not an authority submission, receipt, regulatory approval, or acceptance.
  • The controlled pilot implements the EU baseline. National overlays are a versioned extension point but no national rule package ships in this release.

Related tasks

Your privacy choices

We use essential storage for security and preferences. With your permission, PostHog EU measures filing steps and records a privacy-masked session replay so we can find and fix usability bottlenecks. Replays hide all text, form contents, media, console logs and network contents. This helps us improve SolvencyBridge.

Privacy details