Over the last 18 months JBA has seen a surge of interest in the use of our flood data to support parametric insurance products. With the prospect of more frequent extreme weather events in the future, NGOs, governments, and specialist insurers are looking to simple and affordable products to help people recover from catastrophic events.
What is parametric insurance?
Unlike traditional indemnity insurance, which is triggered by an actual loss sustained, parametric insurance is triggered when an event occurs and meets or exceeds a pre-agreed threshold, i.e. when a parameter or index such as earthquake magnitude, wind speed or flood depth is exceeded. Given that this parameter can be objectively measured, parametric policies can be triggered and settled rapidly and, in some cases, prior to a claim being raised by the insured party. This 'trigger transparency' generally means fewer disputes, which in turn reduces loss settlement costs. As a result, parametric insurance is an attractive alternative in developing sectors and immature insurance markets and is a welcome tool in helping to bridge the protection gap.
While parametric products have only become possible with the availability of high-resolution data and accurate meteorological models, there is still the possibility that the trigger index does not perfectly correlate with the underlying risk. This could result in the client suffering a loss before the parametric insurance is triggered. This is particularly prevalent where exact location of the asset is not known and is referred to as 'the basis risk'. Whilst this basis risk can never be fully eliminated, it can be minimised by employing high-resolution modelling and employing more sophisticated cover such as double trigger thresholds or pay-out structures. In this case the value of the pay-out is increased as the severity of the hazard increases and exceeds multiple pre-agreed thresholds.
How JBA can help you understand and price the risk
The first stage in setting up a parametric insurance product for flood is to understand the probability of flooding and the probability of trigger threshold depths being exceeded at each location. JBA's Global Flood Model can be configured to output location and portfolio level exceedance probability curves for a range of depth thresholds. Clients can choose whether to consider river and surface water flood, or tropical cyclone and non-tropical cyclone events, as well as a range of other parameters tailored to the territory and portfolio. JBA can also combine its Global Flood Maps and Global Flood Event Set to assess the area likely to be inundated.
How to determine if trigger conditions have been met
For earthquake or windstorm events, parameters such as peak ground acceleration or gust velocity can be easily measured and directly related to loss. For flood, however, river flow or rainfall does not simply equate to damage. Rainfall volume and duration, soil and antecedent conditions as well as flood defences all affect the resultant hazard severity. In some markets, sensors have been mounted directly to the exterior of the building to record the depth of water during an event, but for other markets, the cost of sensors may be prohibitive. JBA's Flood Foresight offers a modelled solution. Flood Foresight monitors live river flow conditions and produces high-resolution flood extent and depth footprints taking into consideration the mitigating impact of flood defences. These footprints can then be assessed against the exposure locations to determine whether the trigger thresholds have been met. Flood Foresight can also forecast flooding for up to 10 days in advance allowing pre-event payments to be made.
Recent applications of JBA data
JBA's Global Flood Model and Flood Foresight are currently supporting parametric insurance schemes in Myanmar and Pakistan. JBA's UK Flood Model is currently supporting Flood Flash in the UK.
If you are interested in how JBA can support parametric insurance solutions, please get in touch. You can also find out more about our role in parametric insurance here.
