Peer to Peer Insurance: What are you?
October 5, 2016 Chris Sandilands
There’s much talk about peer to peer insurance (P2P) at the moment – not least because of the launch of Lemonade, a so-called P2P insurer in the US, in September (covered in our Bitesize InsurTech blog post).
But what is this P2P insurance that people talk about?
In lending, P2P is a well-established concept. We would define it as the matching of lenders and borrows in a market, without the involvement of an institutional, risk-bearing balance sheet. Of note for an insurance audience is that this matching does not normally have a “social” element – in other words, if I lend money on Zopa or Ratesetter, I have no control over (or knowledge of) the identity of the borrowers.
(As an aside, note that “purity” of the P2P lending model is being challenged. The FT recently reported that Zopa was securitising loans made on its platform.)
It is close to impossible to do away with the institutional balance sheet in insurance because of unexpected large losses; after all, even the largest insurers buy access to third party risk bearing balance sheets (i.e. reinsurance).
As a result, the definition of P2P insurance is fluid and “P2P insurance” startups have very different propositions. In our opinion, companies described as P2P insurance have just one common attribute: their propositions are designed to increase the transparency of the link between an individual’s claims experience and the premium they pay. In most models – but interestingly least of all in the one we have described as “genuine P2P insurance” – this transparency is combined with incentives for consumers to minimise claims.
In this article, we explore some of the flavours of P2P insurance.
Genuine P2P insurance
Ironically, startups that come closest to providing P2P insurance are not insurance businesses but unregulated companies. Instead, they are structured as risk-sharing syndicates where members agree to indemnify each other against losses. They are explicit about the fact that they are not insurers; the consequence of this is that they cannot guarantee a full indemnity (but the line is, as any insurance professional will notice, blurred).
Teambrella is a Russian startup that is building a genuinely disruptive P2P model, described in a white paper on their website. An interesting hypothesis they put forward it their FAQ section is that “regulation is a tool for resolving conflicts”; Teambrella’s model seeks to remove these conflicts without the need for an arbiter. The proposition could be described as an open source mutual insurance tech platform which allows “teams” to make their own rules (e.g. products, terms and conditions, claims procedures) to, in effect, create mini insurance mutuals on the platform. And they’re doing it all with bitcoin. The model works for any size of risk, in theory – but they note that you might need to buy policies to cover bigger losses when teams are small. Teambrella are truly different – look out for a Bitesize on these guys in the next couple of weeks.
Cycle Syndicate is a London-based non-insurance (i.e. unregulated) startup that “insures” people against stolen bikes. Members join groups and pay an annual membership fee. If a member of the syndicate has their bike stolen, they notify the company who in turn asks other syndicate members if they suspect foul play. If the “claim” is accepted, all members of the syndicate contribute an amount relative to the value of their own bike (a proxy for risk premium). The company retains a 15% fee on the claim. There is no incentive other than one’s own “citizenship” in the syndicate that incentivises a minimisation of claims.
Insurance “micro mutuality”
We define “micro mutuality” as mechanisms whereby sub-sets of risk are underwritten by individuals with minimal institutional balance sheet involvement. This is the closest that insurance startups get to P2P insurance, because insurance companies have to meet all claims and therefore need back-up cover.
Of the companies trading as regulated entities, we think Friendsurance has the model closest to genuine P2P insurance. Friendsurance is a Berlin-based broker offering home contents, private liability, and legal expenses insurance. Customers organise into public or private groups and their premiums are split between a portion that is used to buy insurance and the “cashback pool”. The cashback pool is used to pay small claims and any surplus is distributed to customers at the end of the year (i.e. transparent link between claims and premiums). Should the pot be insufficient, Friendsurance buys stop-loss cover. Large claims are processed through regular insurance coverages. This is “micro-mutuality” because the members of the pool feel like they’re covering each other’s losses and the cashback is driven by claims experience.
Behaviour control
Startups in this group – the biggest – are labelled as P2P but “micro mutuality” is either a secondary element of their proposition, or absent from their business model. Instead, these companies are using group concepts to influence customer behaviour, which means creating incentives for customers not to make false or exaggerated claims. The reward (and the central element of the company’s marketing pitch) is that end-of-year surplus is distributed, normally back to the customer.
Examples:
- So Sure is a London-based MGA offering mobile phone insurance. So Sure encourages customers to connect their friends to their policy. For each friend who buys their own policy and “connects” it to yours, up to a maximum of 8 friends, you get £10 credit. If, at the end of the year, neither you nor any of your connections have claimed, you get up to 8 times £10 back. Whilst the mechanism is technically “micro mutuality”, we have categorised it as behaviour control because its “story” is about getting money back, not about participating in a risk pool. Dylan Bourguignon, So Sure’s founder, is relaxed about not being P2P – he thinks that, if anything, they’re “social insurance”. Look out for a Bitesize very soon.
- Guevara is a London-based startup that originally launched as a broker but is in the process of re-launching as a licenced insurance company. Guevara will allow customers to join either public or private groups (of at least 5 people) to buy motor insurance. Customers pay a premium each year structured as a contribution to a “Protection Pool” and a “fee”. The pool has to be topped up each year but normally won’t need topping up if there are no claims. The fee is payable each year. Guevara is borderline between a micro mutual and behaviour control in our definition. Although the company’s objective is to increase transparency between claims performance and renewal pricing, we think the mathematics are pretty complicated (see their website) and will befuddle most customers. Instead, we think that they will be used by customers who see a general benefit from insurers who reward good behaviour.
- Lemonade is a New York-based startup that claims to be the “world’s first P2P insurance company” (see our Bitesize profile and update). Lemonade takes a 20% “fee” out of all premiums and uses the balance to form a risk pool and also buy reinsurance. Any surplus in the risk pool is distributed to charity at the end of the year; customers are assigned to groups based on which charity they want to support at the end of the year. According to Techcrunch, Lemonade wants to return the money to customers but is prevented from doing so by law – we therefore see it pursuing a similar “behaviour control” model to Guevara and So Sure. 20/14/17 update: Lemonade has now dropped its P2P claims.
None of the above
Finally, we get the impression that any insurance business that has the word “group” in its marketing is put in the P2P box on infographics. They do not have any element of micro-mutuality and do not even try to influence behaviour to reduce claims.
Examples:
- Bought By Many is a London-based broker that builds groups of customers with similar insurance needs. For example, there are groups for pug Insurance and home insurance for cyclists. Bought By Many then works with insurers to build more bespoke products with better pricing. Group buying perhaps, P2P certainly not!
We wrote this blog to highlight the different models that sit under the P2P umbrella. We think any company profiled in this blog could be successful: success is driven by the individual company’s strategy and execution capability, not how it is labelled. Sometimes labels are marketing bluster by the companies, and that’s fine too.
For insurers and brokers there is much to learn about proposition innovation. Whether they are P2P or not, the companies we have covered are all taking a fresh look at the fundamentals of insurance and designing new customer concepts to address them. Established firms would be wise to replicate some of these features in their own propositions.
