The Global Risks Alliance (GRA)
Supervisors see firms mainly through the data firms report to them, and they act on what that data allows them to see. Supervisory data covers regulatory returns, granular loan and transaction records, and the internal risk data that firms aggregate for their own boards. This page explains how data gaps and quality failures hide risk, how reporting standards and identifiers make data comparable, how supervisory technology is changing the work, and who decides what is collected.
Supervisory data · Bank supervision · Cycle 2
| Bank B | High | 3 | 4 | Enhanced |
| Bank D | Medium | 2 | 2 | Standard |
| Insurer Alpha | High | 3 | 4 | Enhanced |
| Fintech F1 | Low | 2 | 0 | Light |
share of firms, thematic review
Fig.Supervisory review console. Design wireframe with placeholder names and illustrative values, not a record or a live system.
Supervisory data is the information firms report to supervisors, from regulatory returns to granular loan and transaction records and the internal risk data behind them. Supervisors act on what the data lets them see, so fragmented definitions, failed aggregation inside firms and reporting lag limit supervision.
How physical and transition risk could reach a loan book under different scenarios. Illustrative design values.
Risk data, quality and reporting. The dossier below sets out how the risk develops, what is measured, where the data come from, how it is modelled and who decides.
The same concept, such as default, exposure or counterparty, is defined differently across templates, regimes and firms. Aggregating across them without reconciliation produces totals that are precise and wrong.
Firms that cannot aggregate risk across legal entities and business lines quickly cannot see their own concentrations. In stress, supervisors asking for an urgent view find the data arrive late, incomplete or inconsistent.
Quarterly returns describe a past position. Liquidity runs, margin calls and market moves unfold in days, so authorities increasingly ask for more frequent and more granular data in stress.
Templates added over decades overlap and duplicate, raising cost without adding insight. Integrated reporting, common data dictionaries and granular collection aim to report once and use many times.
Each term as this page uses it. Where a standard-setting body defines the term, the wording follows that body.
| Term | Definition |
|---|---|
| Regulatory return | A standard report a firm submits to its supervisor on a set template and schedule. |
| AnaCredit | The euro area dataset of loan-by-loan credit data collected by the ECB and national central banks. |
| Data point model | A formal model defining every data item in a reporting framework. |
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 |
|---|---|---|---|---|
| Return resubmission rate | percent of returns resubmitted after validation | quarterly | Supervisory reporting systems | OFFICIAL |
| Validation rule failures | count of failed checks per return | per submission | Taxonomy validation rules | OFFICIAL |
| Risk data aggregation principles adoption | assessment grade per principle for systemic banks | periodic | BCBS progress reports | OFFICIAL |
| Legal entity identifier coverage | percent of reported counterparties with an LEI | quarterly | Granular reporting and GLEIF data | OBSERVED |
| Reporting timeliness | days from reference date to submission | per return | Supervisory reporting calendars | OFFICIAL |
| Granular credit data coverage | percent of loans reported at loan level | annual | Credit register documentation | OFFICIAL |
| Supervisory technology use | presence and type of tools by authority | per survey round | Authority surveys | REPORTED |
The topic as a computational graph: named sources feed the methods, an evidence review gates the outputs, and the decisions stay with the authorities that hold them.
Fig. 02Nodes carry their type, two metric rows and a state; connectors show what feeds what; dashed frames name the owners; the timeline is the cadence. Design of the pipeline, not a record of a run.
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 |
|---|---|---|
| EBA reporting frameworks and data point model | Templates, definitions and validation rules for EU supervisory reporting | Public, reuse with attribution |
| ECB AnaCredit documentation and statistics | Loan-level credit data definitions and aggregate statistics for the euro area | Public, ECB reuse policy with attribution |
| GLEIF Global LEI Index | Legal entity identifiers with reference and ownership data | CC0, public domain |
| FFIEC Central Data Repository | Call reports of US insured banks with definitions | Public domain, US government work |
| BCBS progress reports on risk data aggregation and risk reporting | Assessments of systemic banks against the principles | Public, BIS terms of use with attribution |
| BIS Financial Stability Institute publications on suptech | Surveys and case studies of supervisory technology | Public, BIS terms of use with attribution |
| IMF Financial Soundness Indicators | Harmonised indicators of deposit takers and other sectors with compilation metadata | Public, IMF terms |
Financial Regulation Nexus 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 |
|---|---|---|
| Prudential supervisors and central banks | What to collect, at what granularity and frequency, and how to validate it | Definition maps, burden and duplication evidence, quality practice |
| Statistical authorities and standard setters | Whether to integrate statistical and supervisory reporting into common models | Comparative reporting frameworks and identifier coverage |
| Legislatures | What legal gateways allow authorities to share confidential data | Comparative confidentiality and data-sharing provisions |
| Boards of financial firms | Whether risk data can be aggregated accurately, completely and quickly in stress | Supervisory expectations and assessment practice for risk data aggregation |
Researchers on supervisory data ask whether the data can show the risks that matter.
Supervisory reporting has grown through layers of templates added after successive crises, often with inconsistent definitions. Large institutions struggle to aggregate risk data across legal entities and systems, a weakness exposed during past crises when firms could not quickly report their exposures. Data that is late, inconsistent or incomplete limits supervisors' ability to detect emerging risks.
Reforms aim to standardise definitions, use common identifiers for entities and instruments, and move towards granular data from which authorities can derive what they need. Supervisory technology applies analytics and machine learning to this data. Each step raises questions of governance, confidentiality and the cost of change.
Examining definitions, identifiers and formats with data engineering, interoperability and stewardship. Standards adoption is uneven.
Comparing reporting regimes across jurisdictions. Detailed returns are confidential.
Learning from data failures in past episodes. Public detail is limited.
Using public aggregates. They may hide variation.
The current service level is R0: information and enquiries. 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.