The new buzz around parametric insurance - Part:1 (the what)...

First of all, its not new. Parametric insurance is existing for quite some time. Insurance industry has used this as an avenue to insure the losses occurring due to events which are complex and not covered by traditional products. However, these products are making a comeback in a larger scale. You might wonder why now ? Every industry have seen a metamorphosis lately. For starters, the appetite of risk, behavior and fragmentation of products in the traditional insurance space have put tremendous pressure on the insurance producers to be creative. In addition to that, the quest for new segment, new products and the urge to differentiate from others have made the insurance producers to embrace this. Last but not the lease, the enablement world of mathematical modeling ,computational processing and availability of data in the market place have made it viable.

Simply put, "parametric insurance covers the probability of a event (defined before) happening instead of covering for the actual loss incurred"

In simple terms, if earthquake is an event, the threshold defined is 7 in the Richter scale, and an person or business has taken this policy and has been paying religiously, when this event occurs (when there is earthquake of magnitude 7 or more), agreed upon payout happens irrespective of the damage that occurred exactly for the event to the insurer property.

However its not that simple, often the complexity lies in the details (sounds familiar ?). Insurers get creative in defining scale and payout as a function of various factors and the policy coverage and premium condition itself. However if we look at it from an abstract lens, we can clearly see the attribute of this ecosystem as follows -

Event - The basis of the insurance. Clarity around the event, the nature, the scale and magnitude and the authenticity of definition is very critical.

Pre-Definition - Once agreed during the issuance process, the definition of the event, scale and thresholds are absolutely sacrosanct. They cant change. Its like a contract between stakeholders.

Certifiable Data - The scope of the cycle of insurance, due to the nature of it, depends on the authenticity of data and the declaration of the event. So most of the time, the source of data is agreed and is an industry standard. For example for Earthquake, only the Dept. of meteorology of USA data can be taken as the input.

Automatic/Smooth Payout : This is crux of the existence of such a product. As this product itself is event driven and is not dependent on the indemnification of the loss occurring due to the event, measurement or arbitration of the loss is not a pre condition. This makes the need to pay out to be most of the time fast and automatic.

Risk Model driven quote: This is the basis of product design for an insurance producer. Complex mathematical and simulations are required to determine the quote for an event occurrence. It also requires complex calculation of heuristic and probabilistic determination of likelihood of event. This was one of the struggle for the enterprise before. Unavailability of right talent, speed of computing and complexity of comprehension made it almost impossible. However with advent of explainable machine learning and democratization of the skill aided by exploding computing resources, this problem is possible to be tackled fairly accurately.

Thus far, we have just scratched the definition of parametric insurance and tried to understand what are the component of the ecosystem from the nature of the product prospective. My objective is to run it as an series and explore various product, analyze the pros and cons of them, explore the ecosystem players and compare it to the traditional products in general. In the process, we will touch up the insurtech players and their roles for this kind of products too. Please stay tune. If you want me to focus on a specific aspect of parametric insurance, please do let me know through your comments. Looking forward to publishing the part - 2 soon. Till then, signing off.....