You can judge a risk decision before any capital is committed, but only if you score the reasoning and the limit check, not the profit the simulation happens to produce. A learner can make money in a simulation by luck and lose it by sound judgement, so the outcome alone tells you little. This guide shows how to build a risk management simulation assessment around an explicit rubric, a worked case with a retry, and a new case that checks transfer.
It is written for heads of risk training and for financial crime or compliance leads who have to decide whether someone is ready to take on real exposure. It covers training design only. It is not financial guidance, and every case, limit and criterion should be reviewed by a qualified practitioner and agreed with your own policies.
Quick facts
| Question | Short answer |
|---|---|
| What do you assess? | The quality of the decision and its reasoning, not the simulated profit or loss. |
| What is the yardstick? | An explicit rubric, scored 0 to 2 per criterion, with defined critical errors. |
| What does a good case look like? | Several defensible choices with different loss exposures. |
| How do you check transfer? | A new, unseen case after a delay, scored the same way. |
| What can training not fix? | Unclear limits, poor tools and staffing gaps. |
| Who validates the case and rubric? | A risk or compliance specialist, before use. |
1. Describe the performance problem
Start from the task, not the course. The role must evaluate decisions using an explicit risk rubric: weigh the exposure of each option, check it against the limits in force, record the reasoning and escalate when information or authority is missing.
Here is an illustrative case. A desk is shown three defensible ways to respond to a position that has moved against it. One cuts exposure sharply and accepts a small certain loss. One holds and hedges part of the position. One adds to it on the view that the market will revert. Each is arguable, and each carries a different potential loss. None is labelled "correct" in advance. What separates a ready learner from an unready one is how they choose and how they explain the choice. This is a made-up scenario, not an incident from a customer, and it should stay separate from any real financial guidance.
2. Diagnose the learning and context gaps
Before you build a simulation, compare what the role must do with what current training and assessment test. A product knowledge quiz tests recall. A rubric-scored case tests application and evaluation. Our guide to matching learning activities to Bloom's taxonomy explains how to read that difference.
Then look for causes that training cannot fix:
- Limits. Are the risk limits written clearly, and does everyone read them the same way? The Financial Stability Board's Principles for an Effective Risk Appetite Framework (2013) treat risk limits as one of four areas an institution needs to get right, alongside the framework, the appetite statement and the roles of the board and senior management. If limits are ambiguous, a simulation will only expose that.
- Tools. Can people see their exposure quickly in the real systems?
- Authority. Do learners know when to escalate and to whom?
- Practice. Have they ever made this kind of call with feedback before?
Only the last is a straightforward training gap. Name the others to sponsors and agree who owns each.
3. Design purposeful preparation and practice
A good risk management simulation has three layers, and each has its own job.
Preparation: check the prerequisites. Give a short diagnostic on the rules, limits and concepts the case needs. A recall check confirms prerequisites only. Do not report it as readiness.
Practice: make decisions and get feedback tied to them. Use branching scenarios in which a learner picks a response, gives a reason and sees the consequence. Feedback should refer to the specific decision: "you cited the limit but did not say what you would do if it were breached". Our guide to designing interactive practice with useful feedback goes further.
Review and transfer: discuss one ambiguous choice. Live instructor input adds most where the answer depends on judgement, such as the case where two sound choices pull in opposite directions. Asynchronous review is enough for straightforward checks against a clear criterion. Our guide to connecting self-paced preparation with live coaching shows how to split the two.
4. A worked case with feedback and a retry
Illustrative. A risk or compliance specialist should validate the case and the answer criteria before use.
The case. A position has moved against the desk and is now close to its exposure limit. The learner has three options with different loss exposures, and one piece of market information is missing.
The task. Explain the decision, cite the evidence you used and say what further information or support you need.
A first attempt. "I would hold, because it will probably come back." The reviewer asks what limit applies, how close the position is, and what the learner would do if it went further. The learner has no answer on the last two points.
Feedback and retry. The reviewer points to the limit in the case file and asks the learner to redo the decision, this time stating the exposure after each option.
A stronger attempt. The learner picks a partial hedge, states the resulting exposure against the limit, says what would trigger a further reduction, and asks for the missing data before committing more.
Why each response is defensible or incomplete. Judge against the approved criteria, not against the reviewer's own preference. The first attempt may pick a reasonable option, but it ignores the limit and gives no trigger. The second attempt is right because the reasoning, limit check and escalation are all present, whichever option it picked. Include a variant in which the learner chooses well but never checks the limit, and ask the group how the score should change.
5. Assess capability with an explicit rubric
Score each criterion from 0 to 2. Adapt the criteria with your specialist; this rubric is a proposed starting point, not a validated readiness standard.
| Criterion | 0: absent or unsupported | 1: partial, with omissions | 2: complete and justified |
|---|---|---|---|
| Task execution | No decision, or no record of one. | Decision recorded without its supporting evidence. | Decision and the evidence behind it are recorded. |
| Reasoning | No account of constraints or alternatives. | Some constraints named, uncertainty ignored. | Constraints, alternatives and uncertainty explained, consistently. |
| Adherence to limits | Limit not checked, or breached. | Limit cited but not applied to the exposure. | Exposure checked against the limit for every option. |
| Boundaries | Acts outside approved procedures. | Refers to procedure inconsistently. | Uses approved procedures and seeks help when unsure. |
Define critical errors separately, for example a decision that breaches a hard limit. An overall score must never hide a critical failure. Rubrics help you score the same way each time: a 2007 review of research on scoring rubrics in Educational Research Review looked at their reliability, validity and educational consequences, and reported that reliable scoring can be enhanced by rubrics, especially analytic, topic-specific ones backed by exemplars or rater training. Consistency still depends on agreeing the anchors and checking that two assessors give the same score to the same answer.
6. Plan the transfer
A simulation score tells you what a learner did in the simulation. To learn whether it carries over, test it on a new case after a delay, then add a supervised workplace task.
Two measures fit this design: consistency of reasoning across cases and adherence to risk limits in a case the learner has not seen. Compare a baseline with an unseen follow-up, and define the population, the denominator and the timing before you start. Then check concurrent changes, such as a new limit structure or a change in market conditions, before you credit the training. Our guide to assessing judgement and problem solving in training covers anchors and assessor calibration in more depth.
Where interactive delivery helps
Interactive formats help where people should commit to a decision before they see the group. In a live session, a facilitator can present the case, ask every learner to choose an option and submit a one-line reason, then reveal the spread before discussion. That makes disagreement visible, which is where the most useful teaching happens. The same case can run self-paced for the preparation layer and for the unseen follow-up. AhaSlides supports interactive learning and moving between live and self-paced delivery. Check any specific authoring, feedback or reporting function you plan to rely on against your own workflow, and treat trading simulators, AI and system integrations as external or unconfirmed until you have seen them work.
What training will not fix
- Ambiguous limits. A clear rubric cannot repair unclear limits.
- Tooling. If exposure is hard to see in the real systems, no simulation will change that.
- Staffing and time. People who cannot be released cannot practise.
- Evidence limits. Treat any effect size as context and measure your own population. A simulation score is not proof of performance with real capital.
Frequently asked questions
How do you assess risk management decisions in a simulation?
Score the reasoning and limit check against an explicit rubric, not the simulated profit. Add defined critical errors so a good total cannot hide a serious failure.
What should a simulated risk case contain?
Several defensible choices with different loss exposures, the limits in force, some evidence to cite and at least one gap in the information.
Can a quiz replace a simulation?
No. A recall quiz confirms the rules and limits are known. It does not show whether a learner can apply them to a live-feeling decision.
How do you know the training worked?
Score a new, unseen case after a delay against the same rubric, compare it with a baseline and check what else changed in the meantime.
Who should review the case and rubric?
A risk or compliance specialist who knows your policies and limits, before anyone is scored on them.
Where to start
Pick one decision your role makes that carries real exposure. Write one case with three defensible options, adapt the four-criterion rubric with your specialist, and run it with a small group. Score a new case a month later before you scale it. For a wider view of risk training, see our guide to courses every risk manager needs.
To see how a case like this could run live and be reused self-paced afterwards, explore AhaSlides.
