parg - parametric insurance
Problem: New technologies have the potential to replace incumbents.
- Incumbents are too slow in adapting to changing customer needs because of inflexible legacy IT infrastructure. One example: AI can only be adopted superficially (pricing, simple claims handling, service chatbots).
- Without transitioning into software-first companies, these incumbents are not efficient (administrative cost & claims payments are too high) and customer-focused (speed in adapting to customer needs is too low) enough to attract customers in the mid-term.
- New entrants like Lemonade, WeFox, Getsafe have been forming already, mostly on the promise of being digital-first companies and therefore more efficient and more customer-focused.
Observation 1: price is king
- New entrants struggle to differentiate themselves and will slowly be bogged down with the same level of regulations as incumbent insurers. While some customers prefer a digital-first experience, so far new entrants have failed to establish themselves as service companies that meaningfully compete on non-price parameters. As a result, customers are cost-conscious above all else: lowering the cost structure and predicting the customer’s price sensitivity are as relevant to entrants as to incumbents.
Observation 2: embedded insurance is successful when the timing is right
- Customers do not want to deal with insurance and are happy not to think about it. Therefore, one opportunity to avoid the cost competitiveness spiral is to offer a product to the customer when he/she is open to consider insurance. Simplesurance offers embedded insurance products with success.
Observation 3: AI provides step-function efficiency improvements
- Machine Learning algorithms learn from historical data and leap-frog expert systems. In 2030 McKinsey expects a 50% decrease in manual claims handling compared to 2020 levels (link).
Observation 4: cost to start an insurance carrier goes down with parametric insurance
- Parametric insurance payment is independent of an actual claims event. Therefore, the need for claims review is obsolete. Instead, data providers (oracles) provide verifiable data feeds that trigger payment, e.g. in times of drought. An insurance carrier with a minimal claims process can reduce many auxiliary processes (e.g. processing paper-based input, requesting information).
- If an insurance product is offered on a decentralised platform, the legal guarantee of an insurance contract may be replaced with a technological guarantee. Without a legal guarantee, national entities may not require financial deposits (link), effectively making it possible to start an "insurance-like" company as an UG.
- Etherisk provides a public platform to run smart insurance contracts (link). Arbol is a competitor that raised 7 Mio. EUR in funding in 2021 (link).
Observation 5: Climate change might increase demand for parametric insurance
Potential insurance products
- extreme weather coverage (Deutscher Wetterdienst as oracle)
- potential customers: hotels, restaurants, farmers, car owners
- flight/train delay insurance (Deutsche Bahn as oracle)
- accident insurance (Deutsche Unfallversicherung as oracle)
- care insurance (Deutsche Rentenversicherung as oracle)
- life insurance (institution that issues death certificates as oracle)
Why will it not work?
- market not mature enough yet (ecosystem still being built) --> advantage & disadvantage
- unclear how private data (e.g. death certificates) can be brought on-chain
- too much VC competition --> if recession comes, competition will be reduced
- need successful sales channel --> ?
- need actuarial skill to price products --> ?