GCRI · Systemic Risk Management
Assurance and correction
Assurance asks whether risk management works as described; correction makes sure that when it does not, records and decisions change. Internal audit, independent review, incident investigation and learning systems provide the checks. This page sets out how assurance is organised, why it can miss systemic weaknesses, and how corrections are recorded so that the evidence behind a decision stays traceable after the facts change.
Assurance and correction on the record.
Assurance and correction: how the risk develops.
Evidence, review and methods correction. The dossier below sets out how the risk develops, what is measured, where the data come from, how it is modelled and who decides.
Assurance tests what it is shown
Reviews examine the processes and records put in front of them. Risks that fall outside the scope, or between the scopes of several reviewers, go untested.
Investigations stop at the first cause
Incident reviews that end at the person or component closest to the failure miss the organisational and design conditions behind it. Those conditions remain and produce the next incident.
Corrections must reach the record
When an assumption, dataset or model proves wrong, decisions built on it need revisiting. Without versioned records, the correction is made in one place and the error persists elsewhere.
Actions decay before closure
Audit and incident actions are agreed, then deferred as priorities shift. Overdue actions accumulate and the weaknesses they were meant to fix stay open.
Risk register.
The screen a team would use to follow this subject: its measures, states and decisions. Names are examples and values are illustrative.
Risk register
Assurance and correction · Enterprise A · Review cycle Q4
Residual risk
likelihood × impact, risks per cell
Counts of risks per cell. Shading follows the appetite bands.
Above appetite
| Supplier concentration | Procurement | High | ▲ 2 |
| Data centre flooding | Operations | High | ▲ 0 |
| Model error in pricing | Finance | Medium | ▲ 1 |
| Key staff loss | People | Medium | ▼ 1 |
Risk management process, after the international standard
- Scope, context, criteriaset
- Identifytaxonomy
- Analyselikelihood · impact
- Evaluatevs appetite
- Treatplans
- Monitor and reviewKRIs
- Record and reportboard
Treatment actions
- T-11Dual-source critical componentOwner: procurement · overdue9 d late
- T-14Move backup site above flood levelOwner: operationsdue 14 d
- T-17Independent model validationOwner: financedue 20 d
Assurance checks that risk evidence in every domain can be trusted.
- WaterCouplingWater safety plans are audited against their own hazard analysis.What to watchAudit findings and corrective actions
- HealthCouplingContinuity plans are tested, and the tests are themselves reviewed.What to watchExercise evaluation records
- BiodiversityCouplingRestoration claims need monitoring evidence, not only plans.What to watchMonitoring against baselines
Assurance practice is designed for the Methods and Technical series. The Three Lines Model of 2020 separates management, risk oversight and independent assurance; verification evidence would follow NS-11 of the proposed Nexus Standards.
- Published audit and correction records.
- Lessons where assurance missed a failure.
How assurance and correction is modelled, measured and tested.
The model, its assumptions, its data and its uncertainty, in the form a scientist or engineer would check. Structures follow published methods. The values drawn are illustrative, not findings.
Model, equations and assumptions
The model structure used for assurance and correction, the assumptions it makes and how each is tested.
Ri = f(Hi, Ei, Vi, Ci)hazard, exposure, vulnerability and coping capacityEAL = ∑k pk · Lkover events k with annual probability pkxt+1 = A xt + B ut + εtA carries the couplings, u the shock, ε the error| Assumption | Why it is made | How it is tested |
|---|---|---|
| Hazard and exposure are independent within an event | Keeps the loss integral tractable | Compare with a copula fit on joint records |
| Couplings are linear near the operating point | Lets shocks propagate through A | Perturb beyond the linear range and compare |
| Vulnerability curves transfer between places | Local curves are rarely available | Validate against observed losses where they exist |
- Inputs
- versioned datasets with provenance
- Outputs
- distributions, never single numbers
- Code
- released with the model card and tests
- Reproduction
- same data, code, environment and seed
What is measured, and how well.
Each indicator carries its unit, its cadence, the kind of source behind it and the evidence label a value would carry. No values are shown: a programme supplies them under its own authority.
| Indicator | Unit | Cadence | Source type | Label |
|---|---|---|---|---|
| Principal risks covered by assurance | percent of principal risks reviewed in the cycle | annual | Assurance maps | REPORTED |
| Overdue audit actions | count by age band | quarterly | Audit action tracking | REPORTED |
| Incident investigations reaching organisational factors | percent of investigations | annual | Investigation reports | REPORTED |
| Corrections issued to published risk evidence | count with reason recorded | continuous | Correction log | OBSERVED |
| Repeat incidents | count of incidents with a previously identified cause | annual | Incident databases | OBSERVED |
| Independent reviews of the risk framework | date of last external review | per review | Governance records | DECLARED |
Where the evidence comes from.
Named public datasets and institutional records, each with its licence class. Sources by agreement are used only under their terms and never republished.
| Source | What it gives | Licence class |
|---|---|---|
| Global Internal Audit Standards | Requirements for internal audit independence, planning and reporting, from The Institute of Internal Auditors | Public, IIA terms of use |
| The IIA Three Lines Model | Roles of management, risk functions and internal audit in assurance | Public, IIA terms of use |
| National audit office reports | Independent public sector audit findings on risk and resilience | Public records, licence per office |
| Safety investigation board reports | Investigations that trace organisational and systemic factors | Public records, terms per board |
| Retraction Watch database | Records of retracted and corrected research used in evidence bases | Open, Crossref terms (to check) |
Who decides, and what the evidence serves.
Systemic Risk Management prepares evidence. The decisions below are taken by the authorities named, under their own rules, and never by the platform.
| Who decides | The question | Evidence this page serves |
|---|---|---|
| Audit committees | Whether assurance over principal risks is sufficient and independent | Assurance maps; overdue actions; external review dates |
| Chief audit executives | Where to direct audit effort in the next plan | Coverage gaps; repeat incidents |
| Public accounts committees and legislatures | Whether public bodies have acted on findings | Audit office reports; action status |
Outputs this topic produces
- Assurance maps that show coverage gaps across lines and reviewers
- Investigation protocols that trace organisational and design factors
- Correction and versioning procedures for published risk evidence
- Action tracking templates with ageing and escalation
What assurance experts ask.
- Is the evidence that controls work independent of the people who run them?
- Are models validated against data they were not built on, and are their limits documented?
- When an error is found, is it corrected openly, with a record of what changed and why?
- Do near misses and small failures lead to learning, or are they explained away?
- Who checks the checkers: is there a route for staff and outsiders to raise concerns that bypasses the line being reviewed?
- Are corrections traced through to every decision, report and model that relied on the flawed result?
Why assurance weakens.
Assurance tends to erode quietly. Reviews become routine, findings repeat from one cycle to the next and the people being reviewed learn what reviewers look for. Major failures often follow years of assurance reports that described controls as effective.
Too close to the work
Reviewers who depend on those they review, or who designed the controls, find fewer problems.
Checking the paperwork
Assurance that checks whether processes are documented rather than whether they work.
Unrecorded change
Methods and models changed without records, so no one can tell what earlier results relied on.
How assurance and correction work.
- 01
Model validation
Independent testing of models against data, benchmarks and expert review, using uncertainty quantification and model validation. Limits: validation is harder for models of rare events.
- 02
Internal and external audit
Independent review of controls and governance. Limits: audit samples, and its findings depend on scope.
- 03
Version control and provenance
Records of data, code and method versions behind each result. Limits: requires discipline to maintain.
- 04
Open correction
Published corrections with reasons and dates. Limits: requires a culture that values correction over reputation.
How it is studied, and what the methods cannot say.
- 01Decision analysis under uncertaintyMulti-criteria treatment choice · Decision trees for contingent plansApplied here
- 02Probability and statisticsFrequency and severity modelling · Dependence between risks · Combining data and judgementApplied here
- 03Uncertainty quantification and model validationPropagating input uncertainty · Separating kinds of uncertainty · Deep uncertainty framingApplied here
Where this page’s references lead.
The bodies and documents this page is read against. Status and editions change; each link goes to the issuing body.
Questions and answers.
What does assurance mean in risk management?
Why trace organisational factors in investigations?
How are corrections handled on these pages?
Why publish corrections openly?
Is assurance the same as audit?
Can the platform provide assurance on my risk management?
Contribute evidence from this page.
Reports is the Consortium’s quarterly publication. What you know about this subject can become evidence, a case or a correction in the next edition, credited to you and published under CC BY 4.0 where it is your own work.
- EmailGateway GW-GCRISend it from this pageEvidence, a correction, a case, a method, a review, a translation or profile input. It arrives tagged with this page, so the editor who owns it receives it.
- ZenodoOpen depositDeposit it and submit to the communityPublish your own record on Zenodo with your ORCID iD, then submit it to the Reports community there. Curators review it; the DOI stays yours.
- ReportsGCRISee every way to contributeThe contribution types, how each is reviewed, what credit you receive, the licence and what is never accepted.
Never send personal data, sensitive infrastructure detail or anything you have no right to share. A contribution is reviewed before use; sending it confers no authorship or editorial rights, and sponsorship never buys a finding.
Bring a method, a dataset or a question.
Information and enquiries are open today. An enquiry receives a reference and names the next step; applications, payments, compute allocations and registrations open only when their team, terms and review capacity exist.
Legal documents.
Maps, data and figures on this site are shown for context and general information. They are not advice, forecasts or warnings; the limits of the information are set out in the Disclaimers and Data Notices.