Meeting the challenge of flood reinsurance

Flood risk is an area that presents a growing opportunity for reinsurers, but one with many challenges. Increasing risk and exposure, the continued importance of state-backed insurance schemes, the difficulties in developing efficient catastrophe models and defining what constitutes a flood event are all issues that reinsurers are having to grapple with.

Evolving exposures

Flood exposure is rising relentlessly across the globe. Urbanisation in developed and developing countries has significantly increased values at risk. According to the World Bank, the population living in urban areas has increased from 43 percent to 54 percent in the last quarter century.

Urbanisation, often accompanied by deforestation, profoundly changes ground conditions by increasing impermeable areas. When water flows over concrete and tarmac it does not seep into the soil, dramatically increasing the level of flood risk.

Much recent development has been permitted in areas with significant flood exposure as people’s desire to live near a river or the sea shows no sign of diminishing: 40 percent-plus of the US population lives in coastal counties. Inevitably it seems, investment in drainage and flood defences lags far behind the spend on construction. This is particularly a problem in the many rapidly growing cities of Asia, Africa, and South America, but the developed world is not exempt from this trend either.

In short, this all means that there is now a lot more exposure and a lot more risk. For buyers and sellers of catastrophe reinsurance therefore, flood is undoubtedly a growing focus of attention.

The role of the state

Because flood is effectively a manmade risk, the insurance industry has often argued in the past that flood is uninsurable. This has meant that in many countries, some or all the risk has been assumed by government-sponsored pools or programmes such as the long-established US National Flood Insurance Program (NFIP) or the UK’s rather newer Flood Re. This prevalence of state-backed pools and programmes remains an important feature of the flood insurance market as well as the reinsurance market.

While the aim of most such programmes is to ensure the universal availability of affordable cover, it is difficult to make further generalisations. Flood schemes are all very different, reflecting diverse political systems, insurance markets, and different risk levels. The performance and stability of each programme also vary enormously.

The NFIP perhaps illustrates how it should not be done, but many state-backed pools have successfully balanced providing affordable coverage and the payment of losses over a long period. Those that can achieve this should be excellent reinsurance clients—but what are the key differentiators?

A few key traits mark the best pools from the rest. First is a strong political consensus that a pool is the correct solution. Buy-in from all stakeholders is essential when rate rises or coverage changes are necessary. Without this, the programme may fall apart.

The second essential is a strong emphasis by state authorities on risk mitigation and infrastructure improvement. A clear disadvantage of flat or affordable premiums is their elimination of a key incentive to avoid development on flood-prone land: higher insurance premiums. Strong planning controls and heavy investment in flood prevention and mitigation infrastructure are critical countermeasures.

Finally, it is very important to have well-functioning claims infrastructure. This will support prompt payment and enables essential repairs to be carried out swiftly. Usually the best way to achieve this is to work with private insurers’ own claims adjustment and repair networks. When all these factors are in place, pools make attractive cedants for reinsurers.

Of course, the insurance of flood risk isn’t all in state hands. Driven often by demand from larger commercial customers, the private insurance sector has increasingly developed the skills necessary to insure flood: a good understanding of local conditions together with the application of appropriate policy conditions such as sub-limits and deductibles to price and control flood insurance exposures.

Today they are being assisted by the emergence of high-resolution risk rating tools and exposure management systems that have significantly increased the ability and willingness of insurers to offer this coverage.

Modelling and defining flood events

Reinsurers have increasingly made use of catastrophe models to help price and manage their exposures. However, the development of models for flood perils has generally lagged behind those that model wind and earthquake. To some extent this reflects the major challenges that exist in developing credible flood-loss models.

First, obtaining good quality historic flood data can be an issue. For example, although major floods do not respect national boundaries, statistical collection agencies usually do, hindering efforts to gather comprehensive datasets. Also, recent changes in land use and urbanisation can invalidate much of the historic information obtained.

Flood hazard is very sensitive to changes in ground height, so models must assess this at a very high resolution, which requires a lot of computing power. Still, it will be of limited use if insurers are then unable to generate exposure data at similar levels of resolution. This can be a major processing challenge and increasingly a data protection one too.

Models need also to allow for the impact of flood defences, which can make very big differences to their output. However, accurate information about them can be difficult to obtain.

Finally, modelling flood is, in effect, modelling multiple perils. These include riverine flood, storm surge, and above-ground run-off often interacting with other perils such as wind and rain. Flood events can encompass a large range of different sizes and timings. From sudden events of short duration in a limited area to slowly developing floods affecting a huge area, the variety is considerable.

This complexity makes it hard not only to model but also to draft reinsurance contracts that provide clear definitions of what constitutes an event. The wide variety of event aggregation clauses now in existence reflects the lack of consensus in the market and is yet one more challenge for reinsurers to consider when covering flood exposures.