Peer-to-peer insurance with help of the blockchain technology
The insurance industry is one of the pillars of the global economy. In essence, insurance is a means to cope with risk. In any given situation with a desired outcome, there is uncertainty, and the possibility exists that the outcome will be unfavourable. This possibility of an unfavourable outcome is what we call risk.
Risk is a condition in which there is a possibility of an adverse deviation from a desired outcome that is expected or hoped for. (J. Vaughan)
One may be expecting financial risk, health risk or any other form of risk, and their instinct is to try and hedge it in order to limit the fallout from such a situation. This is where insurance comes in.
The insurance industry helps to hedge or eliminate risks, spreads risks from individuals to the larger community, and provides an important source of long-term finance for both the public and private sectors.
Today it’s almost unimaginable to do business without using some insurance product. Modern insurance companies are among the biggest and wealthiest companies on the planet. Have you heard of Berkshire Hathaway or Allianz maybe? Today’s big insurance companies are efficient and calculated monoliths. The standard model of insurance is that policyholders pay a premium to the insurer company to protect themselves against risk. If the undesired event does not occur, the insurer keeps all the premiums. In case of an unfortunate event, the insurer pays the claim. The model works, there is no doubt about that. However, the house must win. The house must always win in order to have enough to cover everyone’s claims and to make profit. Running an insurance monolith company is much like running a lottery, you win small all the time and occasionally you lose big. Cumulatively, you are bound to win though, especially if you set your math right.
I wonder how things used to work in the past, especially in the pre-state era. Today, both insurance companies and their clients depend on a strong state and the rule of law to force the other party to do as promised.
Manus manum lavat
One of the earliest forms of insurance were death related. Ancient Romans had burial clubs, where members would help cover funeral expenses should another member die. Burials were expensive and would cost a non-trivial amount of money, a situation which could hit hard if you lost your revenue earning family member at the same time you need to cover burial costs.
In the present day, some communities around the Mediterranean Sea still practice similar concepts. If you have ever been to a funeral in the Mediterranean region, especially in rural areas, you may have noticed that people are spontaneously bringing money as a gift/donation to the family of the deceased. This kind of “you help me, I help you” tradition can be understood as a form of peer-to-peer insurance policy. There is no official contract or a club. It is very informal, spontaneous and self-organized. There is no enforcement either, but for sake of science, I’d like to see what happens if one family is spotted to not contribute to several funerals in the row. I wonder would they be excluded from the social contract and lose the support in case of their unfortunate event? As there is no formal agreement, no enforcement and no shared ledger which would transparently show the information to everyone to asses, the entire system depends on the good old grapevine communication channels. However, that is a bottleneck. I do not believe this system can scale up beyond small picturesque Mediterranean villages and small historic towns. It’s a good and noble system, but it cannot scale. Question is how to scale it?
I gave these two historical examples to demonstrate how things used to be done, in p2p fashion and without any sort of enforcement by the organs of the state. Of course, there are examples of how insurance was done in the age of discovery and mercantilism through companies like the Dutch West India Company but let’s not digress. History shows how humans can self-organize into risk-protecting systems if they have the incentive to do so. What we can also conclude is that insurance mechanisms became centralized to allow for trust between parties who inherently don’t trust each other, i.e. systems became centralized to scale up.
Stories from the East
Potential solution comes from the far East, China to be precise. Xiang Hu Bao (相互宝), which literally means “mutual protection,” is a P2P platform launched by Ant Financial in 2018. The app provides a basic health plan to cover costs related to 100 illnesses. It is a form of digitally coordinated health insurance which scales up to 100M users at the time of writing. The platform, which is popular among China’s rural population, is not an insurance product, but a collective claim sharing mechanism built on blockchain technology that offers basic health plans to protect participants against 100 types of critical illnesses. Its 100 million users make up 7 percent of China’s population.
Previous attempts at building a mutual-aid platform in China have failed because of fraud and lack of transparency. By using blockchain technology, Xiang Hu Bao has reduced fraud by making the claim evidence submission process tamper-proof. [Ant Financial]
In the mechanism of XiangHuBao, when insurance claims is filed and confirmed, all the members are obliged to share the total payment of “insurance claims + management fees”. The portion of management fees are received by Alipay. In this case, as the number of claims increases, the profits for insurers increase accordingly, and the conflict of interests between insurers and policyholders is eliminated.
The service offers a maximum one-time payout of 300,000 yuan on each successful claim, covering critical illnesses that include lung, breast and thyroid cancer and critical brain injuries, among others. But is only open to users up to the age of 59. These limits are enforced by the central authority, i.e. the Ant Financial who owns the network. Ant Financial, as the organizer of the scheme, did help it get bootstrap and keep the system balanced but can the system work without a central authority? Can the system work if it’s left unchecked by the central authority and allows the free market to define the max one-time payout, diseases and conditions covered and max-age of the participants?
It is clear that Xiang Hu Bao has brought old P2P concepts back to life by using modern digital technologies which allow the system to scale up to mind-boggling 100M users but it still depends on one single point — one single authority which makes sure that the system works. Xiang Hu Bao is peer-to-peer but it is not self-governed.
Conclusion
The question is whether can we revisit the old concepts by using modern digital technologies. I believe if people are offered a system that allows them to help each other in times of need, people will use it. Learning about Xiang Hu Bao showed me that blockchain can be used to distribute the trust between 100M people, and allow them to self organize. It’s a big social experiment which is challenging the conflict dynamic between policyholders and insurance insurer, where neither party trusts the other.
So why even bother with trying to insert blockchain into this thing? Is it because of the hype? No, it’s not about hype. Blockchain (public blockchain) can be used in this concept to help the system self-organize on a massive scale. It can bring needed transparency, safety of data and easy data verification and bring tools to implement automation. By using tokenization, an economic system can be designed to incentivize positive behaviour and punish bad actors.
A blockchain is a continuously growing list of records, called blocks, which are linked and secured using cryptography. In a digital world, blockchain offers a tool for achieving and maintaining integrity in distributed systems. (Narayanan et al. 2016)
So why not just use Xiang Hu Bao, it’s based on this blockchain thing too? Yes, Xiang Hu Bao is running on proprietary and private blockchain deployed and maintained by Ant Financial. But this is not what I am thinking about, I am thinking about even more widespread system running on a public blockchain which allows user from all across the globe to self organize into a mutual insurance system and enjoy the all the benefits of such a system. In my opinion, having a public blockchain as a backend is mandatory as data safety is mandatory. Also, Xiang Hu Bao is quite China-focused and it kinda depends on one single point of authority to keep the system balanced.
Designing such a system is not a trivial task, there are problems to be handled:
1) KYC on each user is obviously mandatory, as it’s cheap to create new pubkeys.
Anyone can make a private/public key pair. It’s a pretty easy task given proper tools, and thanks to the cryptocurrency boom such tools are abundant and quite user-friendly. Giving you my public key, along with the encrypted data and signature as described is just about as trustworthy as giving you a business card that I had printed for $2. Not that I use business cards, just saying. To really trust that I am who I say I am, and that this is indeed my signature, you need someone to sign off on my identity.
2) How to define the social contract?
When mentioning words “blockchain” and “contract” in the same sentence we usually talk about smart contracts, and most of the discussion around blockchain based smart contracts has been focused on whether they operate in the same way as legal contracts. However, it is argued that most contracts are social in nature and are initiated because the parties trust each other to perform the agreed exchange. Contracts are lawful agreements between parties. In order to create a valid contract, there must be a lawful offer by one party and lawful acceptance by the other party. Contracts usually define the exchange of something of value. To summarize, a contract is a document which needs to answer the following two questions:
- contracts with whom?
- contract about what?
Now should this be readable by a machine or should we leave the interpretation to humans? Or have something in between?
3) How to verify data and events?
Obvious answer is to use oracles, but how to cover each every aspect by using machine oracles? Machine oracles are usable when you need to check if transaction did happen or if A moved to B, digitally. But what about answering is Bob dead? It’s a complex topic. Xiang Hu Bao solves it by allowing all the peers to vote on such questions. Problem with outsourcing the decisions is always tragedy of the commons, i.e. most people will simply not care enough to bother with it. Decision makers must be incentivized.
4) Who is going to make a final decision?
Bob had a life insurance, and now some people are claiming that Bob is dead. If the insurance contract value is a rather large amount of money this is a tricky decision to make. What if they are trying to defraud us? Who is going to make a final call? If the system selects one peer to handle this, how to incentivize him/her/them to do so and again, can we trust this peer to make a right decision?
5) private key management is complicated for many users, loss of keys practically means total loss for the customer
I am not sure how to handle this but via user education and hardware wallets.
References
Narayanan, A., Bonneau, J., Felten, E., Miller, A. and Goldfeder, S. (2016)
Bitcoin and cryptocurrency technologies: a comprehensive introduction. Princeton University Press.
