Climate Risk and Solvency Capital: Capability Gaps vs Regime Gaps

The Bank of England’s March 2023 report on climate risk and solvency capital is one of the more carefully argued pieces of regulatory writing on the topic. Its central distinction — between capability gaps and regime gaps — is useful beyond the UK context because it draws a sharp line that most firms and supervisors blur.

For South African actuaries working under the Solvency Assessment and Management (SAM) regime, the same distinction is the right frame for thinking about how climate risk lands in the standard formula, internal models, and the ORSA.

1. Two Kinds of Gap

flowchart TB
    A[Climate Risk] --> B{Is the risk<br/>captured?}
    B -->|"We cannot measure it"| C[Capability Gap]
    B -->|"The rules do not capture it"| D[Regime Gap]
    
    C --> E[Data, models,<br/>scenario tools]
    D --> F[Standard formula,<br/>time horizon,<br/>risk modules]
    
    E --> G[Firm-level fix]
    F --> H[Regulator-level fix]
    
    style A fill:#0A0F1E,stroke:#00C2CB,color:#F0F0F0
    style C fill:#0A0F1E,stroke:#C9A84C,color:#F0F0F0
    style D fill:#0A0F1E,stroke:#C9A84C,color:#F0F0F0
    style G fill:#0A0F1E,stroke:#00C2CB,color:#F0F0F0
    style H fill:#0A0F1E,stroke:#00C2CB,color:#F0F0F0

The Bank’s framing is that existing capital frameworks — Solvency II, SAM, Basel — already capture climate risk to an extent. Climate-related credit risk flows through the credit module; climate-related catastrophe risk flows through the non-life catastrophe module; transition risk in an investment portfolio flows through the market risk module.

The question is whether the capture is complete. The Bank identifies two reasons it may not be:

• Capability gap. The firm cannot measure the risk to the standard required. The data, the models, or the scenario methodology are not yet mature enough. This is a firm-side problem, fixable with investment.
• Regime gap. The rules themselves do not capture the risk. The standard formula calibrations, the time horizons in the underlying modules, or the way risks are aggregated across modules simply do not contemplate the mechanism by which climate risk crystallises. This is a regulator-side problem, fixable only by updating the regime.

Conflating the two leads to bad decisions. Loading extra capital to compensate for a capability gap is a reasonable interim measure. Loading extra capital to compensate for a regime gap without fixing the underlying rules creates a dependency on firm-level conservatism that will not survive stress.

2. Capability Gaps in Practice

Capability gaps are most visible in scenario analysis and stress testing. For a short-term insurer writing property cover in the Garden Route, Mossel Bay or East London, the core question is how the 1-in-200-year flood or wildfire event will change under a 2°C warming path. That question has three layers:

• Hazard data. Is the climatological model good enough at 10-km spatial resolution and sub-catchment hydrology to project the hazard change? For much of sub-Saharan Africa, it is not, and the downscaling exercise is non-trivial.
• Exposure data. Does the insurer have postal-code level or better location data on the insured property? In many portfolios, aggregation at regional level is the best available.
• Vulnerability data. How does damage scale with hazard intensity for the construction types in the book? The engineering literature is richer for European building stock than for informal or semi-formal South African structures.

These are capability problems. They are solvable with investment in data, third-party providers, and internal actuarial capacity. They do not imply that SAM is inadequate; they imply that the firm has work to do.

3. Regime Gaps in Practice

Regime gaps are harder to diagnose but more consequential. Three candidate regime gaps sit on the watchlist for climate and solvency capital:

• Time horizon. SAM and Solvency II apply a one-year 99.5% VaR horizon to the capital requirement. Climate risk crystallises over multi-decade paths. Annual loss distributions do not carry the forward-looking information about 2040 property concentrations that a supervisor would want the capital framework to reflect. The Bank’s view is that the one-year horizon remains appropriate for capital, and that longer horizons belong in scenario analysis and ORSA rather than in the SCR. That is a defensible position, but it is a position — not a neutral fact.
• Module aggregation. Climate risk crosses traditional risk modules. A transition-driven drop in oil-and-gas equity values is market risk; the credit rating migration of the sovereign that depended on the carbon tax is credit risk; the physical claims event in the coastal property book is non-life catastrophe risk. The standard formula aggregates these through a diversification correlation matrix that was not calibrated against climate-tail co-movement. Whether the correlation assumptions hold under climate-driven stress is an open question.
• Forward-looking vs historical calibration. The non-life catastrophe module in SAM is calibrated against historical event experience. Historical data is, by construction, a sample from a climate that is no longer the operating climate. Pure historical calibration is a regime choice, not an empirical necessity.

Each of these is a potential regime gap. The Bank’s position — and South African PA guidance is consistent with it — is that capability gaps should be closed first, and that regime reform should follow once the evidence base is strong enough to support specific calibration decisions.

4. Implications for SAM

The SAM regime sits in a specific place in the capability-vs-regime discussion.

The standard formula is a calibration exercise. Any regime gap manifests as a mis-calibration — the SCR is either too low or too high for the climate-exposed book. The Prudential Authority’s existing reviews of the standard formula already flag areas where recalibration is expected; climate-driven recalibration is a natural next step, but it is an exercise that takes several iterations.

The internal model route is where capability gaps can be closed most directly. A firm approved for a partial or full internal model can build climate scenarios into its calibration, capture non-linear effects the standard formula cannot, and demonstrate the resulting capital number to the supervisor. This is a high-investment route and is not available to every insurer, but for large multi-line books it is the cleanest path.

The ORSA is where both gap types get their most honest treatment. The ORSA is explicitly forward-looking, explicitly scenario-based, and explicitly requires the firm to form a view on capital adequacy beyond the SCR window. Capability gaps get surfaced in the ORSA as caveats on the scenario outputs; regime gaps get surfaced as the difference between the SCR and the capital the board actually considers adequate.

A SAM first-line function that takes both gap types seriously will produce an ORSA narrative that is much more useful to the board and the PA than one that focuses only on recomputing the standard formula.

5. What to Watch Next

Three items on the regulatory horizon deserve attention:

• PA guidance iterations. The two PA climate guidance notes (GOI 3.1-related) set expectations on governance and disclosure. Further iterations covering scenario standardisation and capital treatment are plausible as international practice consolidates.
• IAIS work on insurance supervision and climate. The International Association of Insurance Supervisors has been active on climate, and its outputs influence how national regulators frame capital expectations over time.
• Bank of England follow-up work. The 2023 report is explicitly a stage-one document. The BoE’s subsequent work on capital frameworks and climate is worth tracking because it tends to set the direction UK-aligned regimes — including SAM — follow.

The capability-regime distinction is not a loophole. It is a discipline. It forces firms to be honest about what they cannot yet measure, and it forces supervisors to be honest about what their rules cannot yet capture. For South African insurers preparing for the next cycle of climate-risk integration into SAM, that honesty is worth more than another provisional capital add-on.