SAM Pillar 1: A Practical Guide to SCR Calculation
The Solvency Capital Requirement is the number regulators, boards, and rating agencies all look at first. Here's what actually feeds it.
South Africa's Solvency Assessment and Management (SAM) regime is built on the same three-pillar structure as Solvency II: Pillar 1 covers quantitative capital requirements, Pillar 2 covers governance and risk management, and Pillar 3 covers disclosure and reporting. Pillar 1 is where the Solvency Capital Requirement (SCR) lives, and it's the single number most boards and executive committees track most closely, because it's the clearest signal of whether the insurer holds enough capital to absorb a genuinely bad year.
SCR vs. MCR: two different thresholds
SAM sets two capital thresholds, and it's easy to conflate them:
- SCR (Solvency Capital Requirement) — the capital needed to absorb losses at a 99.5% confidence level over one year. Breaching it triggers regulatory intervention and a recovery plan, but the insurer is still considered solvent.
- MCR (Minimum Capital Requirement) — a lower, harder floor. Falling below MCR is far more serious and can trigger withdrawal of authorisation.
Most of the day-to-day capital management conversation is about the SCR, since it's the earlier warning signal and the number insurers actively manage against, typically maintaining a buffer well above 100% coverage.
The standard formula's risk modules
Insurers using the standard formula (as opposed to an approved internal model) calculate the SCR by aggregating capital charges across several risk modules, then applying a correlation matrix that reduces the total to reflect diversification — the fact that not all risks are likely to materialise at once. The main modules are:
- Underwriting risk — split by insurance type (life, non-life, health), covering premium and reserve risk, lapse risk, and catastrophe risk. For general insurers, this is usually the single largest component, and it's directly downstream of reserving quality: understated reserves understate this charge too.
- Market risk — exposure to interest rate, equity, property, spread, currency, and concentration risk in the investment portfolio.
- Credit / counterparty default risk — exposure to the insurer's own counterparties, most notably reinsurers. This is where reinsurance recoverables matter: the credit an insurer gets in its SCR for ceded risk depends on the counterparty's credit quality and on the recoverable actually being collectible.
- Operational risk — a capital charge calibrated off volume measures (premiums, technical provisions) rather than modelled directly, intended to capture risks that don't fit neatly into the other modules.
Each module produces its own capital charge; the correlation matrix then combines them into the Basic SCR, before operational risk and any adjustments are layered on top to reach the final SCR.
Diversification cuts both ways. The correlation assumptions that reduce the aggregated SCR below the simple sum of the modules are a regulatory judgment about how risks co-move — not a reflection of how this specific insurer's book actually behaves. Internal model approval exists partly so insurers with genuinely different risk profiles can demonstrate that.
Standard formula vs. internal model
Most insurers use the standard formula — it's prescribed, comparable across the industry, and doesn't require regulatory pre-approval. Larger insurers with risk profiles that diverge meaningfully from the standard formula's assumptions (a concentrated book, an unusual reinsurance structure, unusual correlation between risk types) sometimes pursue a partial or full internal model instead, which can produce a more risk-sensitive — and sometimes materially lower — capital requirement. That comes at the cost of a demanding approval process and an ongoing obligation to validate the model against actual experience.
Where SCR calculations go wrong in practice
- Stale correlation and volume inputs. The standard formula's parameters get updated periodically; running last cycle's factors against this cycle's book overstates or understates real capital needs.
- Manual aggregation across modules. When each risk module is calculated in a separate spreadsheet by a different analyst, the correlation matrix step — where the modules actually get combined — is a common source of quiet errors that don't show up until a reviewer checks the arithmetic by hand.
- Disconnected reserving and capital teams. Because underwriting risk is usually the largest SCR component and is directly driven by reserve estimates, a reserving update that isn't fed promptly into the capital calculation leaves the SCR understating current risk — sometimes for a full reporting cycle.
- Reinsurance credit that doesn't match reality. Claiming full credit for reinsurance recoverables in the counterparty default module, without reconciling against what's actually collectible and aged, overstates capital adequacy.
Keeping SCR current, not just correct
A technically correct SCR calculated once a quarter from a static extract is already out of date by the time it reaches the board pack. The more useful target is an SCR that recalculates automatically as its inputs change — reserves, investment positions, reinsurance recoverables — so a material shift in the book shows up in the capital position within days, not at the next scheduled close.