A risk dashboard is a compact view of selected risk exposures, measures, limits and exceptions, designed to help a particular audience identify what needs attention and investigate it. For a trading book, its practical value is connecting an unusual reading to the positions, assumptions and people behind it before a decision is made.
A dashboard can be a screen, a workbook or a page in a risk report. Its defining feature is the organised view of decision-relevant information, rather than a particular software layout. A dashboard summarises; the supporting records explain.
Two percentages, two questions
Consider a fictional Alpha trading book, reporting in USD, with a snapshot as of 26 October 2026 at 16:00 UTC. All readings below refer to that snapshot.
| Indicator | Snapshot | What it tells the reader |
|---|---|---|
| Value at risk limit utilisation | One-day, 95% VaR of USD 1.2 million; approved limit USD 1 million | The measured metric exceeds its matching limit |
| Open breached limits | 3 distinct breached limits out of 12 monitored limits | The share of monitored limits currently recorded as breached |
The VaR and its risk limit use the same book scope, reporting currency, valuation time, horizon, confidence level and model basis. Their ratio is meaningful because the numerator and denominator match:
The excess is USD 0.2 million, measured in VaR units. It is neither a realised loss nor a maximum possible loss. The approved policy determines what follows the limit breach; a coloured tile alone does not authorise a trade or force liquidation.
For the other indicator, count each monitored limit once. A limit with several unresolved notes still contributes only once to this particular numerator. The VaR limit is one of the three breached limits:
This percentage is a count-based status indicator. It says nothing by itself about the relative severity of the breaches: a small excess and a large excess each count once. It is not a probability of loss, a fraction of capital at risk or a measure of how much the portfolio might lose.
Original synthetic dashboard. The VaR bar uses a 0–150% utilisation scale; the twelve small blocks count limits, with filled blocks denoting breaches. The two percentages have different denominators.
Do not average these percentages into “total risk”. One divides a modelled monetary loss quantile by a monetary limit; the other divides a count of breached limits by a count of monitored limits. A common percent sign does not make their meanings interchangeable. Nor are they independent components: the VaR breach already appears in the count. Any composite score would need an explicit purpose, weighting and interpretation of its own.
What makes the view usable
The headline number needs enough context to survive being copied into a meeting pack. For this example, the useful information travels with the indicator:
- Time and coverage. Show the valuation timestamp and time zone, the book or legal-entity scope, and whether all expected positions and inputs arrived. A refreshed screen can still contain stale prices. Show data cut-offs separately when they differ from valuation time.
- Definition and basis. Carry the units, loss horizon and confidence level where applicable, the model or methodology version, and the limit definition. A trend is comparable only when changes in scope or measurement basis are accounted for.
- Exception detail. Let the reader trace the headline to the individual breached limits, excess amounts, underlying exposures, data-quality qualifications and approved exceptions. A missing input or undefined denominator should appear as unavailable, rather than an apparently reassuring zero.
- Decision and ownership. Identify the risk owner, the action owner where different, the policy escalation route and the next review deadline. An acknowledged alert, an approved temporary exception and a resolved breach are different states.
In the fictional snapshot, the book risk owner reviews the VaR excess and escalates under the approved policy; the exception coordinator checks the action and review deadline for each open breach. Those roles illustrate a decision chain. Actual authority and response rules belong to the institution's mandate.
For a practical continuation, Monitoring and incident handling develops a monitoring contract linking a series and threshold to an owner and action. The full lesson requires membership; its public preview gives the purpose and outline.
A clear display still depends on sound measurement
A dashboard can faithfully display a weak risk model, omit an important exposure, or hide a concentration behind an aggregate. Being within the displayed limits does not establish that every material risk is acceptable. Readers still need the assumptions, stress scenarios and limitations behind the selected measures.
The Basel Committee's banking guidance emphasises timely data, validated reports, appropriate coverage and information suited to the recipient's decisions. These are useful design principles for a dashboard; they do not prescribe one universal screen. See BCBS 239, Principles 5 and 7–10. Refresh frequency and detail should reflect how quickly the risk changes and who can act on it.
Sources
- Basel Committee on Banking Supervision, Principles for effective risk data aggregation and risk reporting (January 2013), Principles 5 and 7–10: timeliness, accuracy, coverage, usefulness and reporting frequency in the banking context.
- John C. Hull, Risk Management and Financial Institutions, fourth edition (2015), §28.1, pp. 577–579: clear risk limits, monitoring and accurate reporting of the risks taken.