Commercial Property Flood Accumulation and Underwriting Appetite

The book is concentrated in one city and one flood zone. The £250m warehouse complex is a single accumulation point, and the many small commercial properties in the same postcode add correlated exposure. The model's 1-in-100-year gross loss of £180m excludes business interruption and contingent business interruption, which can be large for a distribution warehouse: a flood could stop operations for months and disrupt customers and suppliers. The model may also omit off-site utilities, access, and supply-chain dependencies. The per-risk layer of £10m excess of £2m will respond to individual risks, but the catastrophe layer of £150m excess of £50m may be exhausted by a single event, leaving the insurer with significant net loss.

The new £40m warehouse in the same zone would increase the accumulation, and the 25% rate reduction worsens the risk-adjusted return. I would stress the portfolio with a flood event larger than the model's 1-in-100-year scenario, add a BI/CBI load, and check the reinsurance recovery after the per-risk and cat layers. The key issue is not just the modelled loss but the unmodelled and correlated exposure.

I would not accept the new risk at the proposed rate without conditions. The 25% reduction is not supported by the aggregation risk or the unmodelled BI/CBI exposure. The relationship value does not justify writing a loss-making risk that could threaten the cat programme and capital. I would either decline, or offer a much smaller line with a technical price and conditions: a higher deductible, a flood exclusion or sub-limit for BI/CBI, a requirement for flood defences and a business continuity plan, and a maximum aggregate limit for the postcode. I would also check whether the reinsurance programme can absorb the additional exposure and whether the per-risk layer is sufficient.

If the client's group-wide policy requires coverage, I would seek a co-insurance arrangement or a facultative reinsurance quote before binding. The underwriter should be given a clear accumulation limit and a walk-away price. The decision should be documented with the aggregation analysis, because accepting for relationship reasons without pricing the tail risk is not consistent with the insurer's risk appetite.

The insurer needs better data and tools. It should geocode all properties to a finer level than postcode, capture construction, occupancy, flood defences and business interruption values, and link them to a single accumulation system. The flood model should be extended to include BI and CBI, or at least a scenario load should be applied. The insurer should set accumulation limits by zone and by single complex, and monitor them against the reinsurance programme. It should also model the cat layer's exhaustion probability, not just the expected loss, and test the effect of a new large risk before binding.

Governance matters: underwriting guidelines should require a referral for risks above a defined accumulation threshold, and the chief underwriting officer should have authority to decline. The insurer should review the reinsurance structure annually against the updated exposure, and consider parametric or alternative cover for the gap. Finally, it should track near misses and actual flood events to validate the model. These steps would make the aggregation risk visible and controllable, rather than relying on a model that excludes key exposures.