Parametrics: using GIS in insurance

June 5, 2019 09:52


Buying an insurance policy is a form of risk transfer, in that the insurer contractually takes on your risk in exchange for premium. They agree to pay you a sum, up to the limits of the cover you purchase, in the event that something triggers a loss under that policy.

So far, so simple, except that the value of that loss must be proven at the time of making a claim. This produces an inherent conflict of interest between the policyholder and the insurer, since the claimant is incentivised to claim for as much as possible and the insurer to settle for as little as possible, often producing a gap between the actual loss experienced and the indemnification. This process also necessitates human intervention to investigate and negotiate, lengthening in the lifecycle of the claim.

Some risks have traditionally also been difficult to provide coverage for, meaning that individuals and businesses are forced to retain levels of volatility. Examples include compulsory high windstorm deductibles in the US hurricane belt of up to 20% of the insured risk, or business interruption losses where no physical damage is present.

Alternative sources of risk transfer include captives and self-retention, and various contingent forms of risk transfer, traditionally bought by (re)insurers from the capital market (for example catastrophe bonds or other insurance linked securities), but what we are especially excited about are more basic parametric products. These policies pay out a pre-defined sum based on a chosen trigger. Weather triggers are the most obvious, for example if a certain wind speed is reached, or a hail storm is recorded.

The funds are automatically paid to the insured, meaning a frictionless, speedy process. They can use it on whatever they wish, instead of having to prove a certain type of loss that they need the funds for. At MIS, we’ve already helped our clients pay claims from space, and over the past few months I’ve been learning about remote sensing intelligence technologies since joining, and have been thinking of parametric use cases for our intelligence to help build products of the future;

  • Crop and agriculture risks – low yields or failure. We know when this happens because our IoT feeds can record the weather via local stations, hydro gauges, and we can use satellite and aerial data to inform the health of the crop
  • Reduction in output from renewable energy sources, such as wind farms – measured via actual sensors on the infrastructure and inferring loss via local weather patterns
  • Damaged infrastructure disrupting local businesses – Smart cities enables us to sense urban activity at immense scale and in actual real time.

This sort of cover could also lead to better insight for policyholders on what they can do to mitigate their risks, and reduces the risk of fraud since the trigger is pre-set and indisputable. The lower administrative costs for insurers can be passed on in the form of lower premiums, and insight from our data could lead to better predictive modelling, and therefore risk selection.

I love the simplicity and flexibility a parametric product offers to a small business or individual, particularly those traditionally underserved by indemnity based products, and also their potential use in wider, complex programmes to protect risks previously thought to be uninsurable. They are an exciting example of the way that Insurtech is responding to the changing needs of customers. There is huge scope to take advantage of ever more powerful and dynamic data sets to develop innovative products that are designed exactly around customer’s needs.

Vicky Mills, Chief Product Officer


Published on June 5, 2019 09:52

© 2013 - 2019 McKenzie Intelligence Services LTD