Charta is on Google Play: practise on the go. Get the app
Charta

Explain why it is hard to put a robust process for aggregating and reporting risk data in place, and what consequences follow if the data used are of poor quality.

FRM® Part I · Foundations of Risk Management · Principles for Effective Data Aggregation and Risk Reporting · FRM1.FR.7.b

The idea

A robust aggregation process is hard to build because risk data sit in many places. Large banks have grown through acquisitions and run fragmented legacy systems. Data definitions, counterparty identifiers and taxonomies differ across business lines, legal entities and jurisdictions. Much gets stitched together by manual processes and spreadsheets. Ownership of firm-wide data quality is often unclear, because each business optimises its own needs. The investment is costly, slow and competes with revenue projects. The consequences of poor data are practical. Exposures to the same counterparty are missed or double counted, and concentrations go undetected. Reports arrive late or wrong, and decisions are made on a misleading picture. Losses, supervisory action and reputational damage follow. Exam point: poor data quality is a risk in itself, and in 2008 it left banks unable to see their total exposures quickly.

A practice question

From the question bank, with the reason each wrong option is wrong.

Which of the following is a common obstacle to effective risk data aggregation at large banks?

  1. Too little historical market data to calibrate the models used in measuring riskWhy not: Confuses data history for models with the aggregation problem. The obstacle is fragmentation and inconsistency, not sample length.
  2. Fragmented IT systems and inconsistent data definitions across business lines and legal entities - correct
  3. Reliance on a single authoritative source for each risk data element across the groupWhy not: That is a feature of sound architecture, not an obstacle. Its absence is what creates trouble.
  4. Excessive automation of the production of risk reports from the data stores of the bankWhy not: Manual processes and spreadsheets are the weakness. Automation is the cure, not the obstacle.

Growth by acquisition and business-line silos leave legacy systems with different taxonomies, identifiers and definitions, which makes firm-wide aggregation difficult.

A harder one

Which statement best explains why firm-wide risk data aggregation is harder to build than simply buying a larger database?

  1. Risk data are inherently qualitative, so they cannot be expressed in numbers that could be combinedWhy not: Wrongly claims risk data cannot be quantified. The obstacle is consolidation and ownership, not measurability.
  2. Regulators prohibit the exchange of risk data between the legal entities that belong to the same banking groupWhy not: Overstates legal constraints. Cross-border and legal-entity rules can complicate matters but do not ban group aggregation.
  3. Data must be reconciled across entities, systems, definitions and jurisdictions, and no one may own firm-wide data quality - correct
  4. Storage capacity limits mean that no available system can hold the data volumes a large bank producesWhy not: Treats it as a capacity problem. The difficulty lies in integration, definitions, governance and cost, not storage.

Aggregation is a data integration and governance problem: different systems, definitions and jurisdictions must be reconciled and someone must own quality across the group. Buying more storage does not fix that.

The hardest tier

A bank's counterparty exposure report understated its exposure to a corporate group, because subsidiaries were recorded under different names and not linked. Management approved a larger limit on that basis. Which consequence best describes the outcome?

  1. Exposure was overstated, so the bank turned away profitable business, but no loss could possibly result from the data errorWhy not: Wrong direction. Unlinked subsidiaries lead to understated group exposure, and understated exposure can cause loss.
  2. The problem affects reporting only and cannot change losses, since losses depend solely on market movesWhy not: Separates reports from outcomes. Decisions taken on a wrong picture of exposure can enlarge losses.
  3. Regulatory capital rises automatically to compensate for the data weakness, so loss risk is neutralisedWhy not: No such automatic offset exists. Capital is not adjusted to offset data weaknesses in the way suggested.
  4. A concentration went undetected, so the limit rested on an understated exposure and larger losses became possible - correct

Unlinked entities mean group exposure is understated, a concentration is missed and limits are approved on wrong numbers. That is how weak data turn into larger actual losses.

In the app: 6 practice questions on this outcome, spaced review that brings it back before you forget it.

Start free →

More in Principles for Effective Data Aggregation and Risk Reporting