Why ORSA Documentation Matters
An Own Risk and Solvency Assessment (ORSA) is intended to support risk management and decision-making, not merely to produce a report. Good documentation helps people understand how governance operated: who supplied and challenged information, who had authority to decide, what judgments affected the work, and why important choices were made. It also gives management and the board a basis for meaningful review rather than requiring them to reconstruct events from memory.
Documentation does not make a weak decision sound, and a signed document does not by itself demonstrate effective oversight. Its purpose is to preserve clear, relevant evidence of the process. The appropriate format depends on an insurer’s governance arrangements and applicable requirements; there is no single template that is suitable for every organization. This topic is aligned with the CAS Exam 9 (Risk Management for Actuaries) syllabus.
What a Reviewer Needs to Be Able to See
A useful record answers more than “Was the document approved?” It should let a reviewer identify the participants and their roles, the matter considered, the decision-maker, the decision or judgment reached, the principal reasons, and any conditions or follow-up. A concise record can be effective if it answers those questions; a large archive can still be difficult to use if the evidence is scattered or ambiguous.
Keep three ideas distinct. Responsibility describes who performs or owns work. Decision rights describe who may recommend, challenge, approve, or escalate a matter under the organization’s governance arrangements. Evidence shows what happened in a particular instance. A responsibility chart may assign a function, but it does not prove that a particular review or approval occurred. Conversely, a meeting minute may record an approval without clarifying the standing responsibilities of those involved.
| Documentation element | Useful information to record | Question it helps answer |
|---|---|---|
| Roles and responsibilities | Role or named owner, scope of work, expected review or escalation | Who was expected to do what? |
| Decision rights | Who recommends, challenges, approves, or receives escalation | Who had authority to decide? |
| Material judgment record | Issue, options considered, judgment selected, reasons, limitations | Why was this treatment considered appropriate? |
| Decision record | Date, participants, decision, rationale, conditions, actions | What was decided, and on what basis? |
| Review and version evidence | Document version, reviewer comments, resolution, approval evidence | Which materials were reviewed, and what changed? |
The record should be proportionate. A routine administrative choice may need little explanation, while a judgment with a meaningful effect on the assessment or on management’s understanding may need a fuller account. “Material” here means significant enough in the organization’s context to warrant explicit consideration and a record of the reasoning. An organization should state and apply its own criteria consistently rather than relying on an undocumented intuition.
A Practical Record for Decisions and Judgments
A decision record is most useful when it separates the issue, the decision, and the explanation. A short summary of alternatives and limitations can prevent a bare conclusion from being mistaken for a fully considered choice. Where a decision is conditional, the conditions and the person responsible for completing them should be recorded.
State what question was considered, why it required attention, and which version of the relevant materials was used.
Identify who prepared, reviewed, challenged, recommended, and decided, using the organization’s actual decision rights.
Write the outcome in direct terms. Distinguish a recommendation from an approval, and identify any conditions.
Summarize the considerations that drove the choice, relevant alternatives, important limitations, and why the selected approach was considered reasonable.
Assign follow-up owners and dates where needed, and preserve the final version and evidence of review or approval.
A rationale need not reproduce every discussion. It should be specific enough to distinguish the selected option from plausible alternatives. “Management agreed” is not a rationale. “Management retained the existing classification because the available evidence did not support a separate treatment; the limitation will be reviewed when the data source is upgraded” identifies both the reasoning and the caveat.
Worked Examples
Worked Example: Resolving Ambiguous Ownership
A hypothetical insurer is preparing its ORSA. A responsibility chart says the enterprise risk team “owns the assessment.” The finance team prepares several supporting analyses, while a risk committee reviews the package and the board has an approval role under the insurer’s governance arrangements. A reviewer asks who is accountable for coordinating the assessment and who is authorized to approve it. Show how to improve the documentation without changing the stated governance.
Method. Separate the work of coordination and preparation from review and approval. Do not infer approval authority merely from the phrase “owns the assessment.”
Step 1: Identify the ambiguity. “Owns the assessment” does not say whether the enterprise risk team coordinates the process, prepares the entire content, makes decisions, or approves the final package. The finance team’s contribution is also not described.
Step 2: Record the stated activities. The chart can say that the enterprise risk team coordinates the assessment, maintains the process calendar, and consolidates inputs. It can say that finance prepares the specified analyses and identifies relevant limitations in those inputs. These are work responsibilities, not approval rights.
Step 3: Preserve the stated governance roles. Record that the risk committee reviews the package and that the board approves it, as specified in the scenario. The committee’s review should not be described as board approval, and the coordinator should not be presented as the final decision-maker.
Step 4: Specify evidence for this cycle. Retain the version reviewed, the committee’s recorded challenge or comments, how material comments were addressed, and the board’s approval record. Assign an owner and due date for unresolved actions, if any.
Check. The revised documentation answers who coordinates, who prepares a defined input, who reviews, and who approves. It does not claim that a responsibility chart alone proves that review or approval occurred. The common mistake is to treat “owns the assessment” as sufficient evidence of every role; the correction is to state each activity and decision right separately.
Worked Example: Recording a Material Judgment
A hypothetical insurer finds that a third-party data feed used in an ORSA analysis has incomplete records for one business segment. Management considers two treatments: use the feed as supplied with a clear limitation, or delay the assessment until the data provider supplies a revised extract. The risk committee chooses to use the current extract with a limitation and a follow-up review. Draft the essential record.
Method. Record the issue, options, selected treatment, rationale, limitation, and follow-up. The problem supplies no probability distribution or financial parameters, so no quantitative estimate is implied.
Step 1: State the issue. The data feed is incomplete for the specified segment. Identify the feed version and the period or segment affected if those details are available in the organization’s records.
Step 2: State the alternatives. The alternatives considered were using the existing extract with an explicit limitation or delaying the assessment for a revised extract. Recording both makes clear that the choice was considered rather than accidental.
Step 3: State the decision and rationale. The committee chose to proceed using the current extract because, in this hypothetical case, it judged that completing the assessment on schedule with the limitation disclosed was preferable to waiting for a revised feed. This records the stated basis without claiming that the data are complete or that the choice is universally appropriate.
Step 4: Record the limitation and action. State which segment is affected and that conclusions relying on this input should be interpreted with that limitation. Assign an owner to obtain or assess the revised extract, a due date, and a review point for deciding whether the treatment remains appropriate.
Check. The record now explains both the decision and its caveat. A common mistake is to write only “committee accepted the data,” which hides the alternative, reason, and limitation. The correction is to document why the data were used and what will happen next.
Worked Example: Checking Documentation Coverage
For an internal file review, a hypothetical insurer identifies four applicable documentation items for a particular material judgment: (1) a named preparer, (2) a recorded reviewer, (3) a documented rationale, and (4) an assigned follow-up owner. The file contains items 1, 2, and 4, but the rationale is absent. Calculate an internal coverage rate and explain its meaning.
Method. Define an indicator \(I_j\) that equals 1 when applicable item \(j\) is present and 0 when it is absent. The internal coverage rate is the number of present applicable items divided by the number of applicable items. This follows by summing the indicators to count present items, then dividing by the total count.
Step 1: Count applicable items. There are \(n=4\) applicable items.
Step 2: Assign indicators. The preparer, reviewer, and follow-up owner are present, so their indicators are 1. The rationale is absent, so its indicator is 0. Thus \(\sum I_j=1+1+0+1=3\).
Step 3: Calculate. \(C=3/4=0.75\), or 75%.
Interpretation and check. The file contains three of the four specified items, but it lacks the rationale, a substantive gap for a material judgment. The 75% figure is only an internal completeness prompt; it does not establish that the governance was effective or that the judgment was sound. The common mistake is to report the ratio as a passing score and overlook which item is missing. The correction is to inspect the missing item and address the gap directly.
Common Documentation Traps
- Confusing participation with authority. Attending a meeting does not necessarily mean a person had approval rights. Record the role actually performed and the decision rights established by the organization.
- Recording a conclusion without its rationale. “Approved” identifies an outcome, not why the choice was made. Summarize the key considerations and alternatives.
- Using generic ownership language. “The risk function owns ORSA” may obscure the separate responsibilities of coordinators, input providers, reviewers, and decision-makers.
- Leaving limitations outside the decision record. A limitation that affects interpretation should be recorded with the judgment, not hidden in an unrelated file.
- Keeping evidence without making it usable. Minutes, emails, and working papers should be identifiable by date, version, and subject so that reviewers can locate the relevant evidence.
- Treating completeness as quality. A file can contain every expected heading while offering weak analysis. Completeness checks help find omissions; they do not replace judgment or challenge.
When the record is unclear, correct it through the organization’s normal governance process. Do not make a retrospective note appear to have been created or approved earlier than it was. If a later clarification is necessary, identify it as a clarification, state when it was prepared, and preserve the underlying record.
Check Your Understanding
For each scenario, identify the documentation issue or the record that would make governance clearer.
- A hypothetical insurer’s chart assigns the risk team responsibility for “ORSA oversight,” but does not identify who prepares inputs or who approves the final assessment. What two distinctions should the insurer document? (Answer: Separate work responsibilities, such as coordination and input preparation, from review and approval decision rights; identify the role assigned to each.)
- A committee minute says, “The proposed treatment was accepted,” but gives no reason. What minimum additions would make the record more useful? (Answer: Identify the matter, the selected treatment, the principal rationale or alternatives considered, and any material limitation or follow-up action.)
- A material assumption has a named preparer and reviewer, but the file does not say why it was selected over another plausible assumption. What is the key gap, and how should it be addressed? (Answer: The rationale is missing; record the material considerations, the alternative considered, and why the selected assumption was judged appropriate.)
- An internal review lists six applicable documentation items. Five are present, including the decision record, but the missing item is the rationale. Calculate the coverage rate and state why the percentage alone is not a conclusion about quality. (Answer: \(5/6=0.8333\), or 83.33%; the missing rationale may be important, and a coverage percentage does not establish that the decision was well supported.)
- A data limitation was disclosed in a working paper, but the committee decision record does not mention it, and no one is assigned to revisit it. Name two specific improvements to the record. (Answer: Link or summarize the limitation in the decision record, and assign a follow-up owner and review date or trigger.)