How parametric insurance works?

Thanks to Big Data, insurance professionals now have powerful tools to bring their projects to life. They can not only improve existing products, but also design new and more innovative ones.

Parametric insurance has developed thanks to the accumulation of digitized data. It is a direct reflection of this new insurance based on Big Data [4].

The guarantees offered are triggered according to a measurable and objectifiable parameter. They come into play as soon as the agreed upon index reaches or exceeds a defined value. They are most often :

• weather indices: temperatures, precipitation, humidity levels, sunshine levels, wind speeds, etc.
• quantifiable data: airplane flight delays, …

The insurance policy offers lump sums defined upon signature of the contract. Unlike traditional insurance, which is based on the occurrence of a loss that triggers the guarantee, index insurance [5] adopts a completely different operating mode.

The occurrence of a loss is not necessary for the payment of the guarantee. The latter is automatically activated as soon as the triggering threshold of the index defined by the contract is reached or exceeded.

The advantage of index insurance is about its simple operating mode. The insurer automatically pays the lump sum indemnity provided for in the contract. The amount of this indemnity depends on the intensity of the variation in relation to the reference index.

There is therefore no need for an in-depth appraisal to establish the damage: there is no need for an adjuster to establish and estimate the loss, or for the insured to provide proof. The compensation is no longer the result of the prejudice suffered by the insured but of the simple exceeding of an index defined in advance and easy to observe. The measurement of this index is often available from official organizations, hence the following advantages:

• great transparency at the level of the triggering fact,
• unlikely risk of disagreement between the parties,
• quick payment of the insurance indemnity, which is a major element of satisfaction for the insured and also an element of notoriety for the insurer,
• no need for an appraisal,
• low costs because the claim management is less expensive,
• entered information is easily verifiable, hence low risk of fraud and anti-selection.

Example of how parametric insurance works

In 2011, the Caribbean Catastrophe Risk Insurance Facility (CCRIF) set up, in partnership with Swiss Re, a parametric insurance designed to protect farmers in certain Caribbean countries against heavy rainfall.

For Jamaica, the CCRIF has made recourse to the Caribbean Institute of Meteorology and Hydrology (CIMH) and Kinetic Analysis Corporation (KAC) to establish a rainfall simulation model.

With the technical support of Swiss Re and the modeled data, the CCRIF was then able to offer index insurance to its members.

The product developed under the name of XSR (Excess Rainfall) is also based on the resulting data from the Tropical Rainfall Measurement Mission (TRMM).

The TRMM is a joint initiative of NASA and the Japan Aerospace Exploration Agency (JAXA) designed to provide accurate real-time satellite assessment of rainfall aggregated over delineated geographical areas of 625 km2 every three hours.

The map of Jamaica is therefore divided into 28 geographical cells, with a precipitation risk exposure rate being provided for each of the 28 areas (see below):

The precipitation data for each cell is used to calculate a parametric index which is used to determine the following data:

1. Calculation of the cumulative precipitation of the last five days.
2. Triggering of a local event when the aggregate rainfall over five days exceeds 250 mm of rain. The event ends when the cumulative rainfall over five days is below this threshold.
3. Taking into account only the peak of the rainfall aggregated over 5 days allows to measure a loss index. This calculation is obtained thanks to a curve that links the rainfall level to a vulnerability rate.
4. Application of the percentage of exposure of the geographical area to the vulnerability rate of each local event with a view to providing an index of cell loss during the local event.
5. Nationwide loss index estimate is obtained by cumulating the 28 local loss indices day after day. An event is then determined at the national level as a continuous period where the local loss index is positive in at least one of the 28 cells.

⁽¹⁾ A multinational insurance pool covering natural disasters created in 2007 by the Word Bank

Local indices are positive only on days when the peak precipitation is reached or exceeded during a heavy rainfall event. For the remainder of the time, the index is considered as zero.

In the example shown in the diagram above, we see that three events have been reported at the level of the whole country. For the first event, the national index is the sum of the three local indices.

The compensation paid is based on the national indices via the payment function below.