Cell-Captive Insurance in South Africa: The Good, the Bad, and the Untapped Opportunity

How SA’s unique regulatory framework creates both risk and runway for innovation

South Africa remains the global pioneer of third-party cell-captive insurance – a model that lets businesses participate in insurance underwriting without holding a full insurer licence. Capitec Bank, now the country’s largest bank by market capitalisation, recently graduated from a Guardrisk Life cell-captive arrangement into its own licensed insurance operation – proof that the stepping-stone model works. Yet the product category sits at a crossroads. The FSCA’s Conduct Standard 2 of 2022 tightened governance, and a March 2025 alert from Cliffe Dekker Hofmeyr warned that the Prudential Authority’s failure to clarify contractual ring-fencing rules is undermining the very financial stability these structures are supposed to deliver.

The Advantages Worth Defending

Cell captives offer three things traditional insurance cannot easily replicate. First, they provide market access at reduced capital cost. An entrepreneur, retailer, or affinity group can design, price, and distribute tailored insurance products by partnering with a licensed cell-captive insurer such as Guardrisk, Centriq, or Compass, without the multi-year licensing process or the hundreds of millions of rands in solvency capital that a standalone insurer licence demands.

Second, cell owners share directly in underwriting profit. In South Africa, where commissions and fees are capped by regulation, this creates a legitimate economic incentive for intermediaries to manage risk well rather than simply chasing volume.

Third, the structure encourages product innovation. Retailers like clothing chains have used cell captives to sell funeral cover – the single most demanded retail insurance product in South Africa – embedding it at point of sale years before the term “embedded insurance” became a global buzzword.

The Risks That Keep Regulators Awake

The FSCA has identified three structural concerns. The first is conflict of interest: when the cell owner is also the intermediary selling to the public, the incentive to maximise underwriting profit can clash with the duty to act in the client’s best interest. Conduct Standard 2 of 2022 responded by requiring non-mandated intermediary cell owners to be tied agents of no more than one life and one non-life cell-captive insurer.

The second concern is regulatory arbitrage. Cell owners can perform functions similar to underwriting managers but without the same regulatory burden, creating an uneven playing field.

The third – and potentially the most dangerous – is ring-fencing. Under the Insurance Act 18 of 2017, cell structures must be ring-fenced contractually so that one cell’s losses cannot contaminate another. But the legislation does not prescribe how that ring-fencing must be structured, and neither the FSCA nor the Prudential Authority has issued detailed guidance.

Five Practical Recommendations

First, statutory ring-fencing models should be codified. The Prudential Authority should publish approved ring-fencing templates rather than leaving the market to improvise through contractual arrangements that may not survive judicial scrutiny.

Second, parametric and micro-insurance products should be incubated through cells. Cell captives are the ideal sandbox for parametric flood, drought, and load-shedding-surge products.

Third, embedded-insurance APIs should be built into cell agreements from inception. Rather than bolting digital distribution onto legacy structures, new cell arrangements should be designed API-first.

Fourth, annual cell-owner fitness assessments should replace one-time due diligence.

Fifth, data-sharing consortiums among cell-captive insurers should be encouraged. Pooled claims and exposure data would improve pricing accuracy across the entire ecosystem.

The Bottom Line

Cell-captive insurance is not a regulatory loophole; it is a market-access mechanism that has driven genuine product innovation in South Africa for over two decades. But the model will only fulfil its potential if regulators provide the ring-fencing clarity the market needs, and if cell owners invest in the technology and governance that modern policyholders expect.