Illustrative insurance data diagnostics for South African insurers — IFRS 17 contract data fragmentation, claims reserving triangle inconsistency, and SAM return reconciliation failure. What independent advisory finds before major programmes are funded.
These are illustrative scenarios. They reflect patterns found across short-term insurers, life offices, mutual insurers, and financial services groups in South Africa. Names, figures, and identifying details are not drawn from specific clients.
The purpose is to show what an independent data diagnostic examines in an insurance context, what it typically finds, and what it produces as a written output — before core administration platforms are replaced, IFRS 17 remediation programmes are funded, or pricing engine projects are scoped.
For the governance and strategy frame behind these examples, see Insurance Data Strategy.
South African insurers tend to discover data governance gaps at the moment they are most expensive to fix: during an IFRS 17 implementation, an SAM Pillar III review, a pricing model recalibration, or an audit committee challenge on reserve movements.
At that point, the problem is not that data does not exist. The problem is that it exists in multiple versions — claim statuses defined differently across legacy and modern systems, contract groupings maintained in spreadsheets that have drifted from the policy administration record, exposure data that the direct underwriting team and the reinsurance team calculate differently from the same underlying policies.
The diagnostic questions that matter are not technical. They are: who owns each data domain, what definition governs each material field, and when something changes in a source system, who is accountable for updating every downstream calculation that depends on it.
In South Africa, the regulatory and accounting context tightens these requirements. IFRS 17 demands traceable, auditable inputs from contract grouping through to the IFRS 17 subledger. SAM Pillar III QRTs require the same defensibility for prudential returns. POPIA applies across policyholder, intermediary, and claimant data. Insurers that have not addressed data ownership before these obligations land find that addressing them under audit pressure is significantly more expensive.
These diagnostics are relevant to executives at short-term insurers, life offices, and financial groups where:
An insurer with a mix of individual risk products, group schemes, and broker-distributed short-term cover had technically sound IFRS 17 actuarial models and well-documented assumptions. The problem surfaced during an internal control review: contract boundary and grouping logic was applied differently across the policy administration system, the group scheme administrator files, and the manual adjustments maintained by the actuarial team.
The result was not obviously wrong. The IFRS 17 results were directionally plausible. But the insurer could not demonstrate to its auditors — or its own risk committee — that contract groupings were applied consistently across all products and channels, or that the boundary logic used in the calculation matched the contractual terms held in the policy administration system.
Fixing this required ownership decisions, not more modelling: who is accountable for the contract data dictionary, how are group scheme files reconciled to policy records before each calculation run, and who approves changes to boundary assumptions when product terms change.
Actuarial reserving depends on a claims development triangle that reflects actual financial movements on each claim — case estimate revisions, partial payments, recoveries, reopenings. When those movements are recorded inconsistently across legacy and current claims administration systems, the triangle is no longer a reliable view of how the book is developing.
In this scenario, a general insurer found that reopened claims were treated as new notifications in one system and as movements on existing claims in another. Supplier invoices processed through a third-party management platform were not flowing back to the claims system with consistent financial codes. The actuarial team was excluding or adjusting certain claim populations before running the triangle — which meant the reserve was partly dependent on judgements that were not documented as assumptions.
The diagnostic identified three specific data ownership gaps: no agreed definition of “claim finalisation” across systems, no controlled reconciliation between the third-party supplier platform and the claims administration system, and no formal process for locking the data extract used for each actuarial run. Addressing those gaps was upstream of any change to the reserving model itself.
A licensed short-term insurer submitted its SAM QRTs on time each quarter. The numbers were reviewed internally and signed off by the Statutory Actuary. The problem emerged when the CFO’s team compared the capital and solvency metrics in the SAM submission to the figures in the board management pack: the numbers differed on several material lines, and nobody in either team could reconstruct the bridge without going back to the original extracts.
The differences were not errors in either document. They were the result of different extract dates, different filter sets applied to the same source systems, and assumptions about which entities to include in which scope — none of which had been formally documented or agreed between the actuarial, risk, and finance teams.
The diagnostic found that three teams were each maintaining their own version of the capital and solvency data pipeline, with no controlled mapping from source fields to report lines, no locked extract discipline, and no reconciliation between the SAM submission and the management pack as a formal control. The insurer was carrying regulatory and audit risk from a process problem, not a data quality problem.
An insurance data diagnostic does not replace core administration systems, build data warehouses, or retrain actuarial models. It produces:
This is advisory output — not a project plan, not a vendor recommendation, not a technology roadmap. It is the structured thinking that makes those subsequent investments significantly more likely to succeed.