Merger Governance Council Deadlock on Incurred Loss Definition
A specialty insurer completed a merger 14 months ago. Two claims systems remain: Legacy A with 1.2 million open claims and Carrier B with 480,000. The combined governance council has 18 members, meets monthly for 90 minutes, and has approved only 2 of 11 proposed data policies in a year. Business units define 'incurred loss' differently: Finance includes case reserves only; Actuarial adds IBNR; Claims excludes litigation reserves above $250,000. The CDO wants a single enterprise definition before the year-end regulatory report, which is 9 weeks away. The council chair proposes escalating unresolved items to the CIO. Two business unit heads threaten to bypass the council and build local reports. Stewards are part-time, averaging 4 hours a week, and 6 of 10 are behind on issue triage. The council's charter requires consensus, not majority vote, and has no documented exception path.
1. What should the CDO do in the next two weeks to unblock the enterprise definition of incurred loss?
Read the worked answer
The CDO should treat this as a decision-rights problem, not a definition-content problem. An 18-member consensus council is structurally unable to resolve a contested accounting definition in nine weeks; its 2-of-11 throughput confirms the bottleneck.
Actions:
- Create a time-boxed decision team with delegated authority: Finance, Actuarial, Claims, and the CDO. Give it a two-week decision deadline and a named executive sponsor.
- Set explicit decision criteria: regulatory report scope, materiality threshold, reconciliation to the general ledger, auditability, and how local views will bridge to the enterprise figure.
- Adopt a canonical definition with context: for example, gross incurred loss = case reserves + IBNR + litigation reserves above an agreed threshold, with documented exclusions and a reconciliation bridge for Claims.
- Publish a decision record and exception path so business units can use local views only if they disclose the definition used and reconcile to the certified number.
The reasoning is that speed plus legitimacy matters more than a perfect definition. Escalating every item to the CIO would preserve the bottleneck; a delegated decision with a documented trade-off gives the council a reusable pattern for the remaining nine policies.
2. How should the CDO respond to the two business unit heads who threaten to build local reports?
Read the worked answer
The CDO should read the threat as a service failure signal, not primarily as insubordination. The units are bypassing governance because the council cannot decide and because certified data is not available on their timeline.
Response:
- Offer a certified dataset with controlled local flexibility: business units may build local reports, but they must declare which enterprise definition they use and reconcile to the certified figure for regulatory reporting.
- Create a fast exception route with a 10-business-day SLA, owned by the governance office, so unresolved needs do not become shadow reporting.
- Escalate only on misrepresentation risk: if a local report could feed a regulatory submission or external disclosure with a non-certified definition, the CDO escalates with the sponsor.
- Separate authority from service: the council owns cross-unit definitions, but the CDO's team must supply curated data and clear lineage.
This preserves governance credibility while avoiding a confrontation the CDO would likely lose. It also converts the threat into a demand signal that can be measured and prioritized.
3. Which three governance metrics would best show progress after 90 days, and why those three?
Read the worked answer
Use metrics that measure decisions and adoption, not activity:
- Decision cycle time for cross-business-unit data issues - median days from issue intake to decision. Target under 30 days. This directly tests whether the new decision team is relieving the council bottleneck.
- Percentage of regulatory and management reports using a certified definition - target 100% for regulatory and at least 80% for management reports. This measures adoption rather than policy count.
- Steward triage backlog and age - number of issues older than 15 days per steward. With stewards at 4 hours a week and 6 of 10 behind, backlog age reveals whether governance is operationally sustainable.
A secondary metric is exception requests resolved within SLA. Counting approved policies would repeat the original mistake: the council approved 2 of 11 and still had a definition crisis. Throughput, adoption, and backlog age show whether the governance system is actually working.